FR3126078A1 - Active power compensation device with fast switching structure - Google Patents

Abstract

The invention relates to a current compensation device (7) of the parallel active filter type capable of being connected to at least one capacitive energy storage element (3), and to a connection point (C) between an electrical network given (1) and non-linear and linear electrical loads (2), characterized in that the current compensation device (7) has: a power conversion unit (8), comprising at least one power inverter with voltage (9) fast switching; an output filter unit (10), having an output filter (11), connected to the inverter (9), and in parallel to the connection point; the output filter is an L-inductance type first-order filter; a control-command unit (12) comprising a unit for calculating reference currents (25), the reference currents comprising: at least one non-active disturbing current intended to be injected at the connection point (C) in phase opposition , the signal disturbances generated by the nonlinear and linear loads (2), and at the fundamental frequency at least one active current for recharging the capacitive energy storage element (3). Figure 5b

Description

Active power compensation device with fast switching structure

GENERAL TECHNICAL FIELD AND PRIOR ART

The invention relates to the general field of power electronics and its applications on electrical distribution or on-board networks, intelligent buildings and micro-networks, and more particularly to filtering and compensation modules for electrical networks and to injection of power from renewable sources.

It is well known that the quality of the electrical signal delivered by the distribution network has a direct consequence on the performance of the systems supplied by this network, as well as on the life of the electrical equipment which constitutes the network or which is connected to it.

In particular, the harmonic distortion, the phase shift between the voltage and the intensity, i.e. the reactive, as well as the imbalance of the currents and voltages are factors allowing the quality of the energy passing through this network to be expressed.

The term “imbalance” is understood to mean a difference in the physical magnitudes of the signal between the different phases and/or amplitudes, for example voltage levels, intensity.

Harmonic distortion is a measure of signal processing linearity, made by comparing a device's output signal to a sinusoidal input signal.

The non-linearity of the system deforms this sinusoid. The output signal remains a periodic signal which can be analyzed as a sum of sinusoids of frequencies that are multiples of that giving the period, called the fundamental frequency.

Each of these sinusoids is a harmonic of rank equal to the quotient of its frequency by the fundamental frequency. The harmonic distortion rate is the ratio of effective values between the fundamental frequency and the others.

A device comprising non-linear loads connected to an electrical network receives power from the electrical network and reinjects a signal into the network, the signal reinjected into the network being degraded by the operation of the device.

“Non-linear load” means a load based on power electronic components, or others, consuming active power on the supply system (with or without reactive power) and reinjected into the power supply network of the distorting power (related to harmonics). These harmonics can be of conventional ranks (5, 7, 11, 13, etc.) for 3-wire power systems: three-phase non-linear loads (widespread in industrial areas). In the case of single-phase non-linear loads installed in a 4-wire power system (3 phases and the neutral), widely used in residential, commercial or administrative areas, third-order harmonics and their odd multiples (3, 9 , 15, etc.) will circulate in this network in addition to the conventional harmonics.

On the other hand, the phase shift between the voltage and the intensity of the signal implies the appearance of a reactive power which leads, among other things, to the reduction of the active power transmissible in the network.

The development of energy production technologies, particularly from renewable sources, has led to the appearance of many individual production units, particularly in so-called self-consumption or positive energy housing solutions.

When a power production unit, in particular in the case of renewable energies of the solar or wind type, is connected to the network, it is necessary to use one or more inverters or rectifier-inverters, in order to manage and convert the power generated by the unit in direct current and direct voltage in alternating signal, before injecting it into the network.

Inverters most often consist of switches/power electronic components controlled on opening and closing and bidirectional, such as "classic" silicon components such as IGBTs (IGBTs, from English Insulated Gate Bipolar Transistor or Insulated Gate Bipolar Transistor), MOSFETs (MOSFET, from English Metal Oxide Semiconductor Field Effect Transistor ), GTO (GTO, from English Gate Turn-off Thyristor or thyristor with extinction by the trigger) and recently the fast components (with fast switching frequency). The best known of these are power electronics switches made of silicon carbide SiC, (SiC, from English Silicon carbide ) and gallium nitride GaN (GaN, from English Gallium nitride ). By a set of appropriately controlled switchings (usually pulse width modulation), the source is modulated to obtain an AC signal of the desired frequency.

The switching frequencies of the power electronics components of inverters generally induce high-frequency harmonic components in the signal injected into the network, which degrades the quality of the signal passing through the network.

In a set as represented in , comprising a general network 1, a non-linear and/or linear load 2, and a DC energy storage element 3 connected to an inverter 4, it is known to connect at the output of the inverter 4 a filter 5 configured to block the components at high switching frequencies contained in the signal to be injected into the electrical network 1.

Thus, the high frequency harmonics of the signal caused by the switching of the inverter 4, are not transmitted via the filter 5 to the network 1; only the desired signal is generated by the inverter and injected into grid 1 or load 2.

A first-order filter composed of a simple inductor of practically negligible internal resistance (hereinafter called L-Classic), as represented in , in particular does not make it possible to fulfill this function when the voltage inverter used is based on conventional silicon power electronic components (for example IGBT, MOSFET, GTO); the switching frequency of these components, for high switched powers, is industrially limited, for reasons of losses and reliability, to approximately 16 kHz.

Indeed, the higher the inductance of such a filter, the higher the capacity of the filter to prevent the components due to switching from being injected into network 1. However, the higher the inductance, the lower the rate of variation of the intensity passing through the filter and the more the filter will cause a phase shift/delay between the desired real intensity and the intensity to be injected into the assembly. network 1 – load 2 via filter 5.

Conversely, a low value of the inductance allows the majority of the components due to the switching to be injected into the network and consequently to affect the installations and the electrical equipment.

The correct sizing of the first-order output filter will therefore depend on the compromise to be found between the dynamics and the efficiency of the device based on the inverter with conventional silicon components (hereinafter called Ond-Classique), especially when it operates. as an active parallel filter with harmonic depollution.

This compromise is very difficult to fix without the use of an auxiliary passive filter 6 installed at the output of the output filter 5 or upstream of the network side, to filter the high frequency components as shown in .

However, this auxiliary filter 6 can cause unwanted side effects, such as resonance with other passive elements installed on the electrical network 1.

These electrical resonance phenomena in some cases lead to voltage or current peaks, well above the values permitted by the devices connected to the network and cause the destruction of these devices.

This auxiliary filter also causes active power consumption through its damping resistor. In addition, the filtering quality of these auxiliary filters deteriorates over time due to the aging of its passive elements.

A second structure of the filter is a third-order output filter of the LCL type associated with an inverter with conventional silicon components (Ond-Classic). This easy-to-size filter is an alternative to the heavy, bulky, expensive and difficult-to-size first-order (L-Classic) filter. This filter is made up of two inductors (L _f1 , L _f2 whose sum is ten to twenty times less than L-Classic) respective internal resistances (R _f1 , R _f2 ) and a capacitor C _f with a small resistance d damping R _f (see ) whose effect can be overridden by an appropriate command.

It should be noted that (L _s , R _s and e _s ) represent respectively the inductance, the resistance as well as the electromotive force of the electrical network upstream.

This type of filter, thanks to the additional degree of freedom provided by the capacitor C _f , makes it possible to avoid the problems mentioned above in the case of the first-order output filter.

Indeed, the presence of the capacitor C _f makes it possible to reduce the values of the chokes of the two inductors significantly (a ratio of 10 or even more, depending on the switching frequency of Ond-Classic, compared to L-Classic, can be done). In this case, the components due to the switching frequency are blocked on the inverter side, without slowing down the overall dynamics of the system.

However, the inverter must inject both fundamental (reactive and unbalanced) and harmonic components into the electrical network, which implies total control of a very wide frequency bandwidth.

When using the LCL filter associated with linear controllers, a significant phase shift, between the identified disturbing currents caused by the polluting loads, and those compensators injected into the network by the inverter, degrades the quality of harmonic filtering and limits, consequently , the applicability of the active compensator. Indeed, the phase shift increases with the harmonic frequency and with the order of the output filter.

The alternative to overcome this problem of phase shift, encountered when using an LCL filter (connected to the output of an inverter with classic Ond-Classic components) associated with a linear controller, is nonlinear control.

In document WO 2020/007884, the sliding mode control method (SMC for “ Sliding Mode Control” in English) was used, to ensure a desired dynamic response (without phase shift), high robustness/insensitivity to bounded disturbances as well as good control properties over a wide range of operating conditions. However, the classic SMC controller generates a command, at very high frequency (called Chattering ), called discontinuous, applied to the components of the inverter, which can cause overheating which could lead to the destruction of the inverter. In document WO 2020/007884, a conventional SMC (with Sign function) with continuous control was used. Indeed, in order to avoid a discontinuous command, caused by a classic SMC, five controllers were selected by sliding mode:

• the first is associated with a Sign function approximated as a sigmoid function with a second-order sliding surface;
• the second is with a method called AIRD (an artificial increase in relative degree, followed by an integrator);
• in addition, three other nonlinear controllers based on the higher-order sliding mode method are also employed. These are 2-SMC Twisting and 2-SMC super-Twisting second-order sliding mode control algorithm as well as C-HOSMC continuous higher-order sliding mode control algorithm.

Nevertheless, these non-linear controllers are very sophisticated and quite complex to manage and involve the use of a fairly powerful computer. In addition, the control of antiresonance between the LCL filter and the network, despite the solution proposed in document WO 2020/007884, remains a point of which manufacturers are skeptical.

GENERAL PRESENTATION OF THE INVENTION

One of the aims of the invention is to "clean up" the current consumed by a non-linear load, by canceling, on the network side, its harmonic content, the unbalanced content as well as the reactive content of the current, with the aim of improving the quality of the voltage on the grid side. This invention, having a 4-wire structure, is adapted to the harmonic spectrum containing the conventional ranks (5, 7, 11, 13, etc.) of industrial areas as well as those of rank 3 and their odd multiples (3, 9, 15 , etc.) residential, commercial or administrative areas.

Another goal is to maximize the active power producible by a renewable energy generation power unit 100 of Figures 1a, 1b.

Another goal is to optimize the energy consumption of a smart building.

Another goal is to optimize the energy production of a conventional energy production unit (of fossil origin: oil, gas, etc.) within a micro-grid.

Another goal is to optimize the management of energy transiting between production units (renewable and conventional) and consumption units.

The proposed structure is able to avoid:

a- the use of an inverter structure with two voltage levels that is easy to manufacture, control and maintain, but associated with a first-order output filter that is heavy, expensive, bulky and very difficult to design;

b- the use of a new inverter structure other than that with two voltage levels;

c- the problem of phase shift between the disturbing currents identified and those injected into the network due to the LCL type filter associated with a linear controller;

f- the resonance problem caused by the use of the LCL filter;

e- the destructive problem linked to the "chattering" of the inverter when using a conventional non-linear control;

f- the use of an LCL output filter associated with an advanced but complicated to design non-linear control.

The solution proposed in the invention consists in sufficiently increasing the switching frequency of the inverter, thus allowing the use of a simple first-order output filter (with minimal constraints), and consequently in simplifying the method of inverter current control. Indeed, a sufficient increase in the switching frequency of the inverter limits the constraints imposed on the output filter, and a simple inductor thus becomes able to prevent the propagation of the components of the switching frequency of the inverter towards the side. network, without affecting the overall dynamics of the system.

To achieve this objective, inverters based on fast-switching power electronic components (in English “High Switching Frequency”), for example SiC in silicon carbide and GaN in gallium nitride (transistors), are proposed. in the present invention.

Taking advantage of a very high switching frequency, a very low value inductance (a few hundred micro-Henry, for example 200 micro-Henry), can be used to block the harmonic components due to the switching of the inverter without degrading the dynamics of the device.

The output filter used in the present invention is a first-order filter of inductance L type, avoiding resonance between an LCL filter and the network of state-of-the-art devices, and at least 10 times lower, and advantageously at least 20 times lower than the inductance of a first-order filter connected at the output to an inverter with silicon power electronics components of state-of-the-art devices.

Thus, in order to achieve this, while proposing the most reliable, efficient, economical and easy-to-implement structure, the invention proposes, according to a first aspect, a current compensation device of the parallel active filter type, capable of to be connected :

• at its input, downstream of at least one capacitive energy storage element, and
• in parallel, at its output, upstream at a connection point between a given electrical network on the one hand, and non-linear and linear electrical loads on the other.

The current compensation device of the parallel active filter type is characterized in that it has:

• a power conversion unit, comprising at least one power inverter with fast switching voltage structure with silicon carbide (SiC) or gallium nitride (GaN) components (hereinafter called Ond-Cr), the unit converter generating an alternating current, with a frequency band ranging from 50 to 2500Hz, covering the entire frequency band of a non-active disturbing current which presents: all or part of the harmonics, and at the fundamental frequency all or part reactive power and/or current unbalance;
• an output filtering unit, comprising an output filter for each of the phases and a neutral, and connected: on the one hand downstream of the inverter, and on the other hand in parallel to the connection point between the given electrical network and non-linear and linear electrical loads, the output filter being sized to block harmonic components due to inverter switching;

the output filter is a first-order filter of the inductance L type, avoiding resonance between an LCL filter and the given electrical network, with an inductance of less than 1000μH;

• a control-command unit comprising a unit for calculating reference currents, the reference currents comprising:

• at least one non-active disturbing current intended to be injected at the connection point in phase opposition, on the given electrical network side, the signal disturbances generated by the non-linear and linear loads,

exhibiting all or part of the harmonics, and at the fundamental frequency all or part of the reactive current and/or the current unbalance,

• at least one active current for recharging the capacitive energy storage element;

the control-command unit also comprising a switching piloting device which controls the switching of the inverter and which provides closed-loop control of the entire frequency band from 50 to 2500 Hz for injection by the inverter , the non-active disturbance current and the active current, depending on the identification of the reference currents by the calculation unit,

the switching control of the inverter being carried out in such a way as to allow some or all of the non-active disturbing currents injected in phase opposition comprising harmonic currents to pass through the output filtering unit at the connection point, as well as reactive and unbalanced currents at the fundamental frequency, in nonlinear and linear electrical loads,

to satisfy the demand for non-active energy consumption of non-linear and linear electrical loads, while depolluting the given electrical network from these non-active disturbance currents.

Advantageously, the switching frequency of the inverter is greater than or equal to 50 kHz, advantageously greater than 70 kHz, more advantageously greater than 100 kHz.

Advantageously, the first-order filter of inductance L type has a value less than or equal to 400 μH.

Advantageously, the switching frequency of the inverter is greater than or equal to 100 kHz, and the first-order filter of inductance type L has a value less than 200 μH.

Advantageously, the voltage upstream of the inverter is greater than 800 V, and even more advantageously greater than 1000 V.

Advantageously, the device has only two voltage levels.

Advantageously, the device has a so-called digital implementation target processor for the calculations contained in the control command unit, including that of the injected current, having a frequency of less than 100 MHz, a flash memory of less than 2 Mb and a RAM less than 2 Mb.

Advantageously, the control-command unit comprises a calculation unit (linked to the control of the injected current) corresponding to the filter L, less powerful in calculation than the calculation unit of a control-command unit associated with a filter LCL.

Advantageously, the device operates at a fundamental frequency between 40 and 70 Hz. Preferably, at a conventional fundamental frequency between 50 and 60 Hz.

Furthermore, the nominal voltage of the given electrical network according to the present invention is between 180 and 480 V. Preferably, the nominal voltage of the given electrical network is between 230 and 400 V.

Advantageously, the device operates at a fundamental frequency comprised between 360 and 800 Hz, advantageously at 400 Hz and the device operates at a nominal voltage of the given electrical network comprised between 115 and 200 V.

According to a particular embodiment of the invention, the fundamental frequency is between 50 and 60Hz and the nominal voltage of the given electrical network is between 230 and 400V.

Advantageously, the nominal voltage is a phase-to-phase voltage.

Advantageously, the capacitive energy storage element is connected upstream to a renewable energy generation power unit, the conversion unit generating an alternating current, with a frequency band ranging from 50 to 2500 Hz, covering in plus the maximum active current generated by the renewable energy generation power unit, the filter unit passes the active current corresponding to the maximum power point available within the renewable energy generation power unit , to satisfy the active power consumption demand of non-linear and linear electric loads, while ensuring the recharging of the capacitive energy storage element.

Advantageously, the switching control device controls the switching of the inverter, by a controller which is:

- nonlinear by first-order or higher-order continuous sliding mode, or

- linear and takes into account the phase difference between the reference current and the injected current.

In some implementations, the reference currents calculation unit includes an advanced phase-locked loop unit, which extracts, in addition to the angle provided by a conventional phase-locked unit, the amplitude of the direct component of the voltage at the connection point of the given electrical network.

In some implementations, the switching control of the inverter is made such that part of the active current, in the given electrical network devoid of non-active disturbance currents, passes through the output filter unit, when the output of the renewable energy generating power unit is greater than the power consumed by the non-linear and linear electrical loads.

In some implementations, the renewable energy generating power unit is coupled without a chopper or other power electronics devices to the capacitive energy storage element.

In some implementations, the control-command unit comprises:

- a chopper configured to maintain a constant predefined DC voltage at the terminals of the energy storage element of the inverter, independently of the voltage level of the renewable energy source to ensure unchanged harmonic filtering,

- a chopper to manage the charge/discharge of a battery bank or other capacitive energy storage systems,

- a double chopper for the case of an island electrical network.

Advantageously, the control-command unit comprises one of the following two controllers:

- a 1st order continuous sliding mode controller with an approximated sign function in Sigmoid function; Or

- an improved RSTam type linear controller.

Advantageously, the switching control device comprises a pulse-width modulation (PWM) device, in which the command is modulated, in order to make the inverter operate at a fixed switching frequency and adapted to the fast switching of the components. of power electronics constituting the inverter.

In some implementations, the control unit includes a PI type regulator, with an output that represents the maximum power of the renewable energy generation power unit, to regulate the DC voltage of the element capacitive energy storage while ensuring tracking of the maximum active power point of the renewable energy generation power unit, the compensation device comprising a DC voltage regulation loop across the terminals of the capacitive energy storage which supplies the output of the regulator with the maximum power, the voltage across the capacitive energy storage element being equal to the voltage of the maximum power of the energy generating power unit renewable.

Advantageously, the device comprises a unit for calculating reference currents configured to determine the non-active disturbing current flowing in the load, and a unit for calculating reference currents configured to calculate, based solely on the power, the voltage of the maximum active power point of the renewable energy generation power unit, the reference calculation unit delivering, the non-active/disturbance currents of the unit and/or the maximum active current from the unit and of the regulator, which ensures at its output the maximum power of the power unit with renewable energy generation.

Advantageously, the fast switching improves the overall dynamics of the active filter which consequently improves the filtering quality. The upstream voltage (on the Ond-Cr inverter side) reaches higher values, advantageously greater than 800 V (more advantageously, greater than 1000V) against 800 V maximum for conventional Ond-Classic silicon inverters operating, for the same power switched, at a switching frequency of 16 kHz and with an efficiency of 95%. This DC voltage increase improves the dynamics of the active filter and therefore the filtering quality.

Moreover, the first-order output filter is composed of a simple inductor of minimum stresses, subsequently called L-Ond-Cr.

The device according to the present invention has the advantage of being light, economical and very easy to dimension compared to the L-classical filters of the state of the art, in particular thanks to the first-order output filter associated with the undulator. fast switching.

According to another aspect, the invention proposes a current compensation device of the parallel active filter type capable of being connected:

• at its input, downstream of at least one renewable energy generation power unit coupled to a capacitive energy storage element, and
• in parallel, at its output, upstream at a connection point C between, on the one hand, a given electrical network, and, on the other hand, non-linear and linear electrical loads,

the current compensation device having:

• a power conversion unit, comprising at least one fast-switching voltage structure power inverter with silicon carbide SiC or gallium nitride GaN components,

the conversion unit generating an alternating current at the frequency of 50Hz, covering the maximum active current generated by the renewable energy generation power unit;

• an output filtering unit, comprising an output filter for each of the phases and a neutral, and connected: on the one hand downstream of the inverter, and on the other hand in parallel to the connection point C between the electrical network given and the non-linear and linear electrical loads, the output filter being dimensioned to block the harmonic components due to the switching of the inverter;

the output filter is an L-inductance type first-order filter, avoiding resonance between an LCL filter and the given electrical network, and an inductance of less than 1000μH, preferably less than 400μH, more preferably less than 200μH depending on the frequency fast switching;

• a control-command unit comprising a unit for calculating reference currents, the reference currents comprising:

• at least one active current corresponding to a maximum power point of the renewable energy generation power unit;
• at least one active current for recharging the capacitive energy storage element,
• advantageously: at least one current generated by the non-linear and linear loads,

presenting all or part of the reactive current and/or the current unbalance at the fundamental frequency,

the control-command unit also comprising:

- a 1st order continuous sliding mode controller with an approximated sign function in Sigmoid function; or an improved linear controller of the RST _am type.

- a switching control device which controls the switching of the inverter and which provides closed-loop control at the fundamental frequency of 50 Hz for the injection by the inverter of the active current and the non-active disturbing current (reactive and unbalanced), depending on the identification of the reference currents by the calculation unit,

- a PI type regulator, which ensures at its output the maximum power of the renewable energy generation power unit,

the switching control of the inverter being carried out in such a way as to let pass through the output filtering unit at the connection point C, an active current corresponding to a point of maximum power available within the power unit at generation of renewable energy, to satisfy the active energy consumption demand of the non-linear and linear electric loads, while ensuring the recharging of the capacitive energy storage element;

to satisfy the demand for reactive energy consumption of non-linear and linear electrical loads, while decontaminating the given electrical network of these non-active disturbing currents.

Advantageously, the switching control of the inverter being carried out in such a way as to also let pass through the output filtering unit at the connection point C, some or all of the non-active disturbing currents injected in phase opposition comprising reactive and unbalanced currents at the fundamental frequency, in nonlinear and linear electrical loads.

Advantageously, in the compensation device, for an inverter with two voltage levels, the switching frequency of the inverter is greater than 50 kHz, advantageously greater than 60 kHz, advantageously greater than 70 kHz and even more advantageously greater than 80 kHz, and even more advantageously greater than 100 kHz.

Advantageously, the first-order filter of inductance L type has a value of less than 400 μH, and the switching frequency is greater than or equal to 50 kHz.

Advantageously, the voltage upstream of the inverter is greater than 800V, and even more advantageously greater than 1000V.

Advantageously, the control-command unit comprises a calculation unit corresponding to the filter L, less powerful in calculation than the calculation unit of a control-command unit associated with an LCL filter.

Advantageously, the calculation unit is a digital implementation target corresponding to the injected current control unit associated with the filter L.

Advantageously, the renewable energy generation power unit is also coupled without a chopper or other power electronic devices to the capacitive energy storage element.

Advantageously, the control-command unit comprises:

- a chopper configured to maintain a constant predefined DC voltage at the terminals of the energy storage element of the inverter, independently of the voltage level of the renewable energy source to ensure unchanged harmonic filtering,

- a chopper to manage the charge/discharge of a battery bank or other energy storage systems,

- a double chopper for the case of an island electrical network.

According to another aspect, the invention proposes an electrical system comprising a given electrical network, non-linear/linear loads, and a compensation device as presented above.

Optionally but advantageously, the system according to the invention may have the following characteristics, taken alone or in combination:

• the renewable energy generation power production unit is:
  • chosen from the following list: one or more photovoltaic panels, wind turbine(s), fuel cell(s), or other
  • coupled directly to an energy storage element in the case of continuous production, or via an AC/DC power rectifier in the case of alternative production from renewable sources;
• the renewable energy generation power production unit is coupled without a chopper or without other power electronic devices;
• the given electrical network is chosen from the following list: a main electrical network, a local electrical micro-grid that is islanded or connected to the main electrical network, or an on-board electrical network;
• the system further comprises an intelligent building, and in which the control-command unit is connected to a decentralized management unit of the intelligent building, the control-command unit compares the maximum available power of the power unit with generation of renewable energy with the total load of the intelligent building;
• the control-command unit is configured to optimize the consumption of the various devices operating within this intelligent building by distributing the loads corresponding to the non-linear/linear loads according to at least two operating modes:
  • a first distribution mode called adapted consumption mode in which the control-command unit controls the decentralized management unit of the intelligent building so as to adapt the consumption of the intelligent building with the production of the power generation unit renewable energy, so that the total load curve of the intelligent building has a maximum simultaneity factor corresponding to the operation of all the useful loads of the building at the same time, within the limit of the renewable energy produced,
  • in the event of insufficient renewable energy production, a second mode of distribution called modulated consumption mode in which the control-command unit controls the decentralized management unit of the intelligent building so as to modulate the consumption of the devices of the intelligent building to tend to a substantially constant intelligent building total load curve as a function of time;

advantageously, the first mode of distribution takes priority;

advantageously, in the second distribution mode, the total load curve of the smart building is adapted to the generation state of the given electrical network;

• the system also includes:
  • a local network connected to the given electrical network, and
  • power generation units of conventional origin, and
  • a semi-decentralized management system, and
  • a plurality of power units with renewable energy generation connected to the local network, by a compensation device, each compensation device being connected to the semi-decentralized management system to which it communicates information relating to the current energy production and to come, of each of the power production units with renewable energy generation, and
  • a plurality of consumption stations corresponding to the non-linear/linear loads, each of the consumption stations being connected to the local network and equipped with a compensation device, connected to the semi-decentralized management system to which it communicates information concerning the instantaneous consumption and future consumption according to the programmed operation of consumption items;
  • a plurality of power units generating renewable energy and linear and non-linear loads of positive energy smart buildings, which ensure self-consumption and where any excess energy is stored or exchanged with the other smart buildings or delivered to the local network via the control-command unit of the compensation device, in coordination with the semi-decentralized management system in the event of an exchange with the local network;
• the semi-decentralized management system is also configured to ensure the economic distribution of production from conventional energy production units;
• the semi-decentralized management system is also configured to intervene when the total energy demand within the local network is greater than the total production, with the decentralized control units of the intelligent buildings, via the control-command units of the compensation devices, to make them switch to a modulated consumption mode;
• the semi-decentralized management system is also configured to control the distribution on the local network of the power coming from the given electrical network if the estimated total production does not cover the demand.

PRESENTATION OF FIGURES

Other characteristics and advantages of the invention will emerge from the following description, which is purely illustrative and not limiting, and must be read in conjunction with the appended figures in which:

(state of the art) is a diagram representing an electrical network supplying a load and a renewable energy generation power unit connected to an inverter, via one or more energy storage elements, comprising a filter according to prior art;

(state of the art) represents this same network, with another prior art filter structure;

(state of the art) is a diagram representing an electrical network supplying a load with an inverter, according to the state of the art, with silicon power electronic components associated with an LCL output filter;

(state of the art) is a diagram representing an electrical network supplying a load with an inverter, according to the state of the art, with silicon power electronic components associated with an L-Classic output filter;

(included in an active compensation device according to the invention) is a diagram representing an electrical network supplying a load with an inverter, according to the invention, with fast switching silicon carbide power electronic components associated with a filter L-Ond-Cr output with minimum constraints;

(state of the art) is a diagram representing the structure of a conventional active compensation device (with inverter, according to the state of the art, of silicon power electronic components associated with an LCL output filter or L-Classic), in particular the structure of the control-command unit;

(according to an implementation of the invention) is a diagram representing the structure of an active compensation device (with an inverter, according to the invention, of fast switching silicon carbide power electronic components associated with a L-Ond-Cr output filter with minimum constraints), in particular the structure of the control-command unit;

is a diagram representing the advanced PLL phase locked system according to the invention;

(state of the art: conventional inverter and LCL) is a diagram representing an electrical system according to the state of the art, in which the compensation device is inserted, in parallel, between a renewable energy production unit , within a self-consumption building, and the network-loads assembly;

(inverter, according to the invention, with fast switching and L (L-Ond-Cr) with minimal constraints) is a diagram representing an electrical system, according to an embodiment of the invention, in which the compensation device is inserts, in parallel, between a renewable energy production unit, within a self-consumption building, and the network-loads assembly;

(state of the art: conventional inverter and LCL) is a diagram representing an electrical system according to the state of the art, in which the system is connected to a positive energy intelligent building;

(inverter, according to the invention with fast switching and L-Ond-Cr with minimal constraints) is a diagram representing an electrical system according to an embodiment of the invention, in which the system is connected to a positive energy smart building ;

(state of the art) represents the regulation loop of the energy storage element with the objective, among others, of extracting the maximum active power from the renewable energy generation power unit;

(according to an implementation of the invention) represents the regulation loop of the energy storage element with the objective with the advanced PLL system, among other things, of extracting the maximum active power from the unit power to renewable energy generation;

(state of the art) represents the integration of the method for calculating this maximum active power in the algorithm for identifying/calculating non-active currents (eg harmonics, reactive and unbalanced);

(according to an implementation of the invention) represents the integration of the method for calculating this maximum active power in the algorithm for identifying/calculating non-active currents (eg harmonics, reactive and unbalanced) by using the PLL advanced ; the identified output currents include the neutral wire current;

(state of the art) is a diagram representing an electrical system according to the state of the art (based on conventional inverter and LCL), in which the electrical system includes a semi-centralized management system which manages the production conventional energy from the network as well as, in case of emergencies, the mode of operation of the centralized units via the active compensation devices connected, in turn, to the renewable energy production units;

(according to an implementation of the invention) is a diagram representing an electrical system according to an embodiment of the invention (based on fast switching inverter in silicon carbide and L-Ond-Cr of minimum constraints) , in which the electrical system includes a semi-centralized management system that manages the conventional energy production of the network as well as, in case of emergencies, the operating mode of the centralized units via the active compensation devices connected, to in turn, to renewable energy production units;

shows a general structure of the parallel active filter according to an implementation of the invention;

(state of the art) represents a phase-equivalent diagram of the LCL third-order output filter according to the state of the art;

represents an equivalent diagram per phase of the output filter of L-Classic according to the state of the art or of L-Ond-Cr with minimal constraints according to an implementation of the invention;

(state of the art) represents the gain diagram of the LCL third order output filter according to the state of the art;

shows the gain diagram of the L-Ond-Cr first order output filter with minimal constraints according to an implementation of the invention;

(state of the art: Conventional inverter & LCL with an RST controller) represents, according to the state of the art, a general block diagram of the current control algorithm of the active filter;

(state of the art: Classical & L-classical inverter with a Proportional Integral (PI) controller) represents, according to the state of the art, a general block diagram of the current control algorithm of the active filter;

(according to an implementation of the invention: Fast-switching inverter & L-Ond-Cr of minimum stresses with an RSTam or continuous SMC controller with sigmoid function) represents, according to an implementation of the invention, a block diagram general active filter current control algorithm;

represents an effect of the phase shift on the quality of compensation of the active filter; and more particularly the intensity of the load;

represents an effect of the phase shift on the quality of compensation of the active filter; and more particularly the setpoint current and the injected current;

represents an effect of the phase shift on the quality of compensation of the active filter; and more particularly the network current with the effect of the phase shift as well as the estimated ideal current without the effect of the phase shift;

(state of the art: an LCL output filter) represents the gain and phase diagrams of the transfer function of the control loop of the active filter with RST which is a linear controller based on a placement of the poles of the closed loop (control loop);

(according to an implementation of the invention: an L-Ond-Cr output filter with minimal constraints) represents the gain and phase diagrams of the transfer function of the control loop of the active filter with RST which is a linear controller based on a placement of the poles and RSTam, according to an implementation of the invention, which is based, in addition, on the placement of the zeros, of the closed loop (control loop);

(state of the art) represents tracking and single-phase control signal with sliding mode controllers using Sign, Sigmoid and artificial relative degree increase control functions;

(state of the art) represents single-phase tracking and control signal with SMC, C-HOSM, 2-SMC Twisting, 2-SMC Super-Twisting control algorithms and Lyapunov approach;

(according to an implementation of the invention) represents the single-phase tracking and control signal with the Sign function discontinuous sliding mode and Sigmoid function continuous sliding mode controllers as well as the RSTam enhanced pole placement controller;

represents the harmonic spectrum (due to chopping) of the grid side current for a fast switching inverter & an L-Ond-Cr output filter with minimal constraints (L-Ond-Cr =200 µH and F-Ond-Cr =100000kHz );

represents the harmonic spectrum (due to chopping) of the mains side current for a classic inverter & an L-classic output filter (L-Classic =2mH and F-Ond-Classic= 16kHz);

represents the harmonic spectrum (due to chopping) of the current on the mains side for a classic inverter & an output filter L with minimal constraints (L-Ond-Cr = 200mH and F-Ond-Classic =16 kHz);

represents the harmonic spectrum of the current on the load side;

represents the simulation for the case of a continuous SMC controller with sigmoid function associated with a fast switching inverter and L with minimum constraints 200 µH: the temporal analysis before and after filtering of the network current Is, of the identified currents Iref and injected Iinj superimposed as well as the THD of the current Is;

represents the simulation for the case of a continuous SMC controller with sigmoid function associated with a fast switching inverter and L with minimum constraints 200 μH: spectral analysis after filtering of the network current Is;

represents the simulation for the case of an RSTam controller associated with a fast switching inverter and L with minimum constraints of 200µH: the time analysis before and after filtering of the network current Is, of the identified currents Iref and injected Iinj superimposed as well as of the THD of the current Is;

represents the simulation for the case of an RSTam controller associated with a fast switching inverter and L with minimum constraints of 200µH: spectral analysis after filtering of the network current Is;

represents the simulation for the case of a continuous SMC controller with sigmoid function associated with a classic inverter with a switching frequency of 16 kHz and L with minimum constraints of 200 μH: the temporal analysis before and after filtering of the network current Is, of the identified currents Iref and injected Iinj superimposed as well as the THD of the current Is;

represents the simulation for the case of a continuous SMC controller with sigmoid function associated with a conventional inverter with a switching frequency of 16 kHz and L with minimum stresses of 200 μH: the spectral analysis after filtering the network current Is (components due to the hash);

represents the current-voltage and power-voltage characteristics of a photovoltaic generator;

represents a DC voltage regulation loop of the unit of the storage element of the inverter;

(state of the art) represents an algorithm for identifying (non-active) disturbing currents integrating the tracking of the maximum power point;

represents an algorithm for identifying (non-active) disturbance currents integrating maximum power point tracking with, according to an implementation of the invention, the advanced PLL system;

shows the control scheme of the advanced PLL system, according to an implementation of the invention;

represents the characteristics of the photovoltaic generator (PVG) for different illuminations and temperatures;

represents the time analysis of the voltages of the maximum power Vmp of the PV generator and the capacitor of the inverter Vdc;

represents the active and reactive power balance of the Active Filter-GPV-Network-Load assembly.

DESCRIPTION OF ONE OR MORE MODES OF IMPLEMENTATION AND REALIZATION

The invention relates to electrical systems shown in figures 2-3, comprising a given electrical network 1 with 3 phases and a neutral, a load 2 non-linear or linear or both, connected to the given electrical network 1; a capacitive energy storage element 3; and a compensation device 7 connected on the one hand at its input, downstream, to the capacitive storage element 3 and on the other hand in parallel at its output, upstream to a connection point C located between the electrical network given 1 and load 2.

The current compensation device 7 of the active filter type is able to be connected:

• at its input, downstream of at least one capacitive energy storage element 3, and
• in parallel, at its output, upstream to a connection point C between on the one hand a given electrical network 1, and on the other hand non-linear and linear electrical loads 2.

The compensation device 7 has:

• a power conversion unit 8, comprising at least one fast-switching voltage structure power inverter 9 with components made of silicon carbide (SiC) or gallium nitride (GaN), (or other materials of technologies of fast power electronics), the conversion unit 8 generating an alternating current, with a frequency band ranging from 50 to 2500 Hz, covering the entire frequency band of a non-active disturbing current which presents: all or part harmonics, and at the fundamental frequency all or part of the reactive power and/or the current unbalance;
• an output filtering unit 10, comprising an output filter 11 for each of the phases and a neutral, and connected: on the one hand downstream of the inverter 9, and on the other hand in parallel to the connection point C between the given electrical network 1 and the non-linear and linear electrical loads 2, the output filter 11 being sized to block the harmonic components due to the switching of the inverter 9,

the output filter 11 is a first-order filter of inductance L type, avoiding resonance between an LCL filter and the given electrical network 1 and an inductance of less than 600 μH;

• a control-command unit 12 comprising a unit 25 for calculating reference currents, the reference currents comprising:

• at least one non-active disturbing current intended to be injected at the connection point C in phase opposition, on the given electrical network side 1, the signal disturbances generated by the non-linear and linear loads 2,

exhibiting all or part of the harmonics, and at the fundamental frequency all or part of the reactive current and/or the current unbalance,

• at least one active current for recharging the capacitive energy storage element 3;

the control-command unit 12 also comprising a switching piloting device 21 which controls the switching of the inverter 9 and which provides closed-loop control of the entire frequency band from 50 to 2500 Hz for the injection by the inverter, of the non-active disturbing current and of the active current, according to the identification of the reference currents by the calculation unit 25,

the control-command unit 12 requiring a calculation unit corresponding to a filter L;

the switching control of the inverter 9 being carried out in such a way as to allow some or all of the non-active disturbing currents injected in phase opposition including harmonic currents to pass through the output filtering unit 10 at the connection point C , as well as reactive and unbalanced currents at the fundamental frequency, in the 2 electrical nonlinear and linear loads,

to satisfy the demand for non-active energy consumption of non-linear and linear electrical loads 2, while depolluting the given electrical network 1 from these non-active disturbance currents.

Advantageously, the calculation unit of the control-command unit 12 is adapted to the control unit 23 associated with a filter L.

Advantageously, the switching control of the inverter 9 and the output filtering unit 10 at the connection point C form a compensator against disturbing currents.

Advantageously, the device has a so-called digital implementation target processor for the calculations, having a frequency of less than 100 MHz, a flash memory of less than 2 Mb and a RAM of less than 2 Mb.

The control-command unit 12 comprises a PI-type regulator 62, with an output which represents the power delivered via the given electrical network, necessary to regulate the DC voltage of the capacitive energy storage element 3 around a reference voltage greater than 800 V.

Advantageously, the compensation device 7 is equipped with a fast-switching inverter 9 Ond-Cr and an L-type output filter 11 Ond-Cr, which can operate, according to the reference currents supplied by the calculation unit 25, as a parallel active filter, to ensure the filtering of all or part of the harmonics, while offering the possibility of compensating all or part of the reactive power and/or unbalanced current.

Advantageously, the inductance of the output filter 11 is less than 1000 μH, advantageously between 10 μH and 800 μH, the inductance of the output filter 11 is such that the inductance is at least 10 times lower than the conventional inductance a first-order filter connected at the output to an inverter with conventional silicon components.

Thus, the inductance can be between 50 and 1000 μH, advantageously between 100 and 600 μH, more advantageously between 100 and 400 μH, even more advantageously between 100 and 200 μH depending on the fast switching frequency greater than 50 kHz, advantageously greater than 70kHz, even more advantageously greater than 100kHz.

The higher the switching frequency, the smaller the inductance can be, within the physical and technological limit of the components.

For example, the output filter inductance can be chosen from the following list: 50μH, 55μH, 60μH, 65μH, 70μH, 75μH, 80μH, 85μH, 90μH, 95μH, 100μH, 105μH, 110μH, 115μ5μH, 1.12μH, 1.12μH 130μH, 135μH, 140μH, 145μH, 150μH, 155μH, 160μH, 165μH, 170μH, 175μH, 180μH, 185μH, 190μH, 195μH, 200μH, 205μH, 210μH, 215μH, 220μH, 225μH, 230μH, 235μH, 245μH 255μH, 260μH, 265μH, 270μH, 275μH, 280μH, 285μH, 290μH, 295μH, 300μH, 305μH, 310μH, 315μH, 320μH, 325μH, 330μH, 335μH, 340μH, 345μH, 350μH, 365μH, 365μH 380μH, 385μH, 390μH, 395μH, 400μH, 405μH, 410μH, 415μH, 420μH, 425μH, 430μH, 435μH, 440μH, 445μH, 450μH, 455μH, 460μH, 465μH, 470μH, 480μH 505μH, 510μH, 515μH, 520μH, 525μH, 530μH, 535μH, 540μH, 545μH, 550μH, 555μH, 560μH, 565μH, 570μH, 575μH, 580μH, 585μH, 590μH, 595μH, 600μH 705 μH, 710μH, 715μH, 720μH, 725μH, 730μH, 735μH, 740μH, 745μH, 750μH, 755μH, 760μH, 765μH, 770μH, 775μH, 780μH, 785μH, 790μH, 795μH, 800μH, 815μH 830μH, 835μH, 840μH, 845μH, 850μH, 855μH, 860μH, 865μH, 870μH, 875μH, 880μH, 885μH, 890μH, 895μH, 900μH, 905μH, 915μH, 915μH, 920μH, 925μH, 940μh, 955μH, 960μH, 965μH, 970μH, 975μH, 980μH, 985μH, 990μH, 995μH, 1000μH.

The compensation device 7 can be connected to a system having a renewable energy generation power unit 100.

Advantageously, the renewable energy generation power unit 100 is connected downstream to the capacitive energy storage element 3, that is to say that the renewable energy generation power unit 100 is located upstream in the structure, and that the output (downstream) of the renewable energy generation power unit 100 is connected to the input of the capacitive energy storage element 3.

The command control unit 12 is further configured to detect the maximum power point MPPT ( Maximum Power Point Tracking ) of a renewable energy generation power unit 100 and to generate, via the inverter 9 , the active current corresponding to this maximum power; the voltage V _dc across the terminals of the capacitive energy storage element 3 is equal, in this case, to the voltage of the maximum power of the renewable energy generation power unit 100.

The conversion unit 8 generates alternating current, with a frequency band ranging from 50 to 2500 Hz, additionally covering the maximum active current generated by the renewable energy generation power unit 100.

The filter unit 10 passes the active current corresponding to the maximum power point available within the renewable energy generation power unit 100, to satisfy the demand for active energy consumption of the non-linear electrical loads 2 and linear, while ensuring the recharging of the capacitive energy storage element 3.

The renewable energy generation power unit 100 is coupled without a chopper or other power electronic devices to the capacitive energy storage element 3.

In one embodiment, the device operates as a parallel active filter, the device in this case only injects non-active currents for harmonic filtering, with the possibility of compensating for reactive power and current imbalance.

In this embodiment, operation as a renewable energy system may be an option.

In this embodiment, advantageously, a charge/discharge management option for a battery bank or other energy storage systems can be provided.

In one possible embodiment, the device provides the two previous operating modes: injecting the maximum power of the renewable energy generation power unit 100, while providing harmonic filtering with the possibility of compensating the reactive power and /or current imbalance.

In this embodiment, a charge/discharge management option for a battery bank or other energy storage systems can be provided.

The 8 Power Conversion Unit

The power conversion unit 8 comprises, as represented in FIGS. 2c and 3b, the Ond-Cr inverter 9 with rapid switching with 3 phases and a neutral which has a structure with two voltage levels.

The 9 Ond-Cr inverter is based on SiC or GaN components or other technologies with a switching frequency greater than 50 kHz; advantageously higher than 80 kHz (depending on the technology of the components and its evolution), to gain in dynamics (in speed), which makes it possible to achieve high efficiency for a much less bulky inverter with good heat evacuation as well as better reliability, compared to an inverter with conventional Ond-Classic silicon components. Moreover, the fast switching improves the global dynamics of the active filter which improves, consequently, the quality of filtering. In addition, the use of fast switching components makes it possible to raise the voltage on the DC side of the inverter to more than 800 V; advantageously greater than 1000 V, which also improves the dynamic range of the device and consequently the filtering quality.

The inverter 9, with 4 wires and 4 arms, comprises for each of the 4 arms, two switching devices 13 and 14 each connected on the one hand to an output terminal of the capacitive energy storage element 3 and on the other hand to a common connection point A to which one of the 4 wires is also connected.

This connection point A located, for each arm, between the first switching device 13 and the second switching device 14 forms a connection point for one of the 4 wires of the inverter 9 connected to the other end of the unit filter 10 whose output is connected to connection point C.

Each switching device 13, 14 is composed of a power electronics switch D, controllable on opening and on closing (MOSFET SiC, HEMT in GaN or others) with or without conventional diode or with fast switching in antiparallel B (bi-directional operation).

The inverter 9 may have a transistor structure (SiC or GaN) making it possible to achieve high efficiency for an inverter that is much less bulky, with good heat removal as well as better reliability in comparison with an inverter with conventional silicon components (Ond-Classic: IGBT, MOSFET, GTO).

The inverter 9 emits, on each wire, a signal having an intensity I _inj .

The inverter 9 is configured to inject currents which cover a wide frequency band, usually between the frequency of the fundamental, for example 50 Hz, up to the harmonic order 50, having in the example a fundamental of 50 Hz a frequency of 2500 Hz.

Advantageously, the power conversion unit 8 has a fast component structure with diodes in antiparallel.

The power conversion unit 8 may not have diodes in antiparallel.

The control-command unit 12

The control-command unit 12 is configured to control the inverter 9 so that the inverter 9 generates a signal of intensity I _inj configured in phase opposition to cancel, at the connection point C, the non-active disturbances of the signal generated by the load 2 and injected into the given electrical network 1.

The control-command unit 12 is further configured to compensate for the imbalance of the signal passing through the non-linear or linear load 2.

Imbalance is understood to mean a different current consumption (in amplitude and/or in phase) depending on the different phases of several single-phase 2 loads.

Advantageously, the control-command unit 12 is configured to compensate for the reactive power consumed in the non-linear or linear load 2.

Advantageously, the control-command unit 12 is configured to filter/compensate for the harmonics of the signal transiting on the given electrical network side 1.

The control-command unit 12 is further configured to regulate (charge and maintain constant) the voltage V _dc across the terminals of the capacitive energy storage element 3. The voltage V _dc , thanks to the use of the fast component Ond-Cr inverter, can be higher than 1000V, which promotes the overall dynamics of the active filter. While for the same sizing power as the Ond-Cr, industrial limitations are imposed on the classic silicon inverter, in switching frequency at 16 kHz and in DC voltage at 800 V for efficiency between 95- 97%.

The control-command unit 12 can include one of the following two controllers:

- a 1 ^st order continuous sliding mode controller with an approximated sign function in Sigmoid function; Or

- an improved linear controller of the RST _am type (RST _am controller).

Advantageously, the RST _am controller performs a placement of the poles in a closed loop, via the polynomials R and S, to ensure, among other things, the usual tracking, in amplitude, but also a placement of the zeros to ensure the tracking in phase. Indeed, for the classic RST controller, as well as all linear controllers, the problem of phase shift between the identified harmonics and those injected by the device 7 degrades the tracking aspect between its signals, in terms of phase, and consequently prevents the applicability of device 7 for harmonic filtering.

Advantageously, the control-command unit 12 comprises a PI type regulator 62, with an output which represents the maximum power of the renewable energy generation power unit 100, to regulate the DC voltage of the capacitive energy storage 3, while ensuring tracking of the maximum active power point of the renewable energy generation power unit 100. The compensation device 7 comprising a DC voltage regulation loop across the terminals of the capacitive energy storage element 3, which supplies the output of the regulator 62 with the maximum power, the voltage across the terminals of the capacitive energy storage element 3 being equal to the voltage of the maximum power of the unit of power to renewable energy generation 100.

Advantageously, the control-command unit 12 comprises:

- a chopper, configured to maintain a constant predefined DC voltage across the terminals of the capacitive energy storage element 3 of the inverter 9, independently of the voltage level of the renewable energy generation power unit 100 for ensuring unchanged harmonic filtering;

- a chopper, to manage the charge/discharge of a battery bank or other energy storage systems;

- a double chopper, for the case of an island electrical network.

The control-command unit 12 is configured to regulate the voltage across the terminals of the capacitive energy storage element 3 as well as to identify and control the current injected, via the filter unit 10, at the connection points C for the 4 wires and to control the inverter 9 to generate this current. The identification of the currents to be injected at the connection points C is done by the calculation unit 25 of the reference currents.

Advantageously, the 4 wires (the three phases as well as the neutral) are controlled in a similar and independent manner by the control-command unit 12, which therefore comprises a control-command chain per wire.

Advantageously, to control the switching devices 13, 14 of an arm/wire, the control-command unit 12 comprises two outputs, one connected to the trip pole of the first switching device 13, the other being connected to the tripping pole of the second switching unit 14. Arm denotes a part of the switching devices 13, 14 comprising at least one terminal serving as an entry or exit point for the electrical circuit.

The two outputs can be connected jointly to the output of a logic comparator 18, so as to simultaneously control the switching of the two switching devices 13, 14.

Advantageously, one of the outputs of the control-command unit 12 comprises a logic level inverter 19, so that the activation of one of the switching devices 13, 14 leads to the deactivation of the other, being able to be during a single control pulse thanks to their common connection, upstream of the logic level inverter 19, to the same comparator 18 logic.

The logic comparator 18 emits a logic output signal corresponding to the comparison between an output signal of a saturation element 20 and an output signal of a switching driver device 21.

The switching pilot device 21

Advantageously, the switching control device 21 controls the switching of the inverter 9, by a controller 23 which is:

• nonlinear by first-order or higher-order continuous sliding mode, or
• linear and takes into account the phase difference between the reference current and the injected current.

The switching control device 21 may include a pulse width modulation (PWM) device, in which the command is modulated, in order to operate the inverter 9 at a fixed switching frequency and adapted to the rapid switching of the power electronic components making up the inverter 9.

Advantageously, the switching control of the inverter 9 is carried out such that part of the active current, in the given electrical network 1 devoid of inactive disturbing currents, passes through the output filtering unit 10, when the production of the renewable energy generation power unit 100 is greater than the power consumed by the non-linear and linear electrical loads 2.

The switching driver device 21 emits a signal (called a carrier wave) at a very high frequency (above 80 kHz initially) predetermined, depending on the dimensioning of the components, so as to cause the switching of the switching devices 13, 14.

In most cases, control laws, designed to control voltage inverters that are connected to the given power grid 1 through an LCL filter, were originally established for renewable energy systems (photovoltaic and wind) .

However, when operating as a parallel active FAP filter, a resonance problem at the LCL cutoff frequency, as described on the , a phase shift problem caused by linear controllers 23, as described in the , associated with the LCL filter, between the identified signal (I _ref : supplied by the calculation unit 25 of the reference currents) and that injected by the inverter 9 (I _inj ), as described in FIG. 14, the problem of chattering for the classical discontinuous nonlinear control of the SMC type, as well as the complexity of the implementation of nonlinear controllers with continuous control, involving the use of a powerful calculation unit, constrain the use of the LCL filter.

A chattering problem is when the classic SMC controller generates a command at a very high frequency.

Advantageously, the device to be controlled is based on an Ond-Cr inverter connected in parallel in the middle between the given electrical network 1 and the linear and non-linear loads via an L-Ond-Cr inductance type filter.

The controller 23 can be chosen from these two types of controllers:

the first controller is linear of the RST _am improved pole placement type;

- the second is a continuous sliding mode type nonlinear controller, as described in the .

The classic RST controller (which provides only amplitude tracking) is based on pole placement, while the improved RST _am method performs pole and zero placement, which allows the error in amplitude and in direction to be minimized as much as possible. phase between the identified signal/current and the injected one, as described on the .

The RST controller is composed of three polynomials R, S and T, advantageously, these polynomials are of first order corresponding to the order of the output filter L-Ond-Cr. The polynomial T is a simple gain which makes it possible to make the error in amplitude, between the input and output signals of the control loop, tend to zero for the entire band of harmonic frequencies.

The improved RST _am controller preserves the same R and S polynomials while placing zeros in the control loop to further minimize the phase error between the input and output signals.

In both embodiments, the controller 23 performs a control which makes it possible to ensure a desired dynamic response, a strong robustness in stability and in performance with a very good rate of rejection of disturbances in a wide range of operating conditions.

For the sliding mode controller, it is known that a classic sliding mode control with a so-called Sign function, generates a very high frequency control of the inverter switches (called Chattering ) (discontinuous control) , to ensure finite-time convergence to the slip surface where the system states are subsequently maintained, even in the presence of bounded perturbations.

The sign function of a signal is given by the relation:

In practice, the voltage inverter is controlled by a fixed or variable frequency limited switching function. Switching at very high frequencies causes overheating which can lead to the destruction of the inverter.

In order to avoid operation in discontinuous control mode, controller 23 advantageously operates with a sliding mode with continuous control associated with an approximated sign function as a sigmoid function.

This method consists in replacing the Sign function by a continuous approximation. It is a Sign function approximated to a Sigmoid function given by the formula:

with the thickness of the vicinity of the sliding surface.

The system, in this case, no longer converges towards the desired value, but towards a neighborhood of the latter, which makes it possible to limit switching at very high frequencies.

It should be noted that the sliding mode controller with sigmoid function, used according to one aspect of the invention, is much lighter and is easily realizable compared to that used for the LCL filter. Indeed, the relatively high (three) order of the LCL filter implies a minimal sliding surface second-order, which involves deriving the current tracking error twice :

with

According to one aspect of the invention, the output filter is a simple L-Ond-Cr minimum stress inductor, so the sliding surface is only the current tracking error without any shunt. . The control is in this case much easier to design, to manage and no longer requires a numerical target of powerful implementation.

A digital implementation target designates an electronic component, such as a processor, a microprocessor, a microcontroller, an FPGA, the calculation unit, etc., which can perform tasks and calculations, thus the component is capable to execute the algorithms of the control-command part of the proposed device.

Tasks can be alarm scanning, ADC control (analog and digital convention), component driver control, PWM/PWM signal generation driver control, etc.

The calculations can be the execution of the calculation unit algorithms (the identification of disturbing currents), the detection algorithm of the MPPT, the algorithm of the advanced PLL, the calculations of the regulation loop of the voltage of the capacitive energy storage element 3, calculations of the injected current regulation loop, etc.

This controller makes it possible to avoid a discontinuity of the command, in particular caused by variable commutations of the inverter at very high chattering frequencies, caused by a classic sliding mode control.

It should be noted that the active filter structure based on a classic silicon inverter associated with a state-of-the-art L-Classic inductor is controlled by a proportional-integral (PI) linear PI type control, as described in the .

Advantageously, the switching control device 21 comprises a pulse width modulation device, in which the command is compared with a carrier wave at a switching frequency corresponding to the rapid switching of the Ond-Cr predetermined according to the sizing the components of the compensation device 7.

This allows the Ond-Cr 9 inverter to operate at a fixed switching frequency (above 80 kHz initially).

As the switching frequency of the Ond-Cr inverter is increased up to 80 kHz (or even more), the 11 L-Ond-Cr filter will have a synergistic effect thereby improving the blocking of high frequency components, due to switches/hashes.

One of the aspects of the invention is to associate with the Ond-Cr inverter 9 an output filter 11 of the L-Ond-Cr inductance type which is light, not bulky, economical and very easy to size, compared to the filter L -Classic associated with an inverter with classic silicon components (Ond-Classic: IGBT, GTO) of the state of the art.

Moreover, this inductance, whose inductance has a size equal to or less than the sum of the inductances (L _f1 +L _f2 ) of the LCL filter, described for example on the ), ensured a retention or blocking effect of the components due to the switching frequency very close to that of the LCL filter, without risk of resonance and with a control much less complicated to implement, and for a less powerful calculation unit .

The controller 23 is adapted to the filter used, in the case of the state-of-the-art LCL filter, the controller is very sophisticated and difficult to design, so a more powerful calculation unit is required.

In the present invention, on the contrary, the controller associated with the L-Ond-Cr filter is lighter and simpler, consequently the necessary calculation unit is less powerful.

Advantageously, the digital implementation target is the processor which makes it possible to perform all the calculations of the device, this means that all the algorithms, unit 25, controllers 23, 62, 26, etc. of the control command part 12 are executed by the implementation numeric target.

The digital implementation target may have a frequency lower than 100 MHz, a flash memory lower than 2 Mb, for example a few hundred kb and a RAM lower than 2 Mb, for example a few tens of kb.

For example :

1- Digital implementation target linked to the L-Ond-CR filter and lightweight controller of the present invention:

in this example of the invention, the device needs a target that has the following characteristics:

- single-core,

- frequency of the order of 30MHz,

- 32 kB flash memory,

- 8 KB RAM,

- it does not contain FPU Floating Point Unit.

2- Digital implementation target linked to the LCL filter and sophisticated controller, as used in the state of the art:

In this example of the state of the art, we need a target that has the following characteristics:

- dual core + DSP acceleration,

- 1GHz frequency,

- 16 Mb flash memory,

- 32 Mb RAM,

- with FPU.

In addition, the fast component inverter technology allows the voltage on the inverter side to reach high values (more than 1000 V), compared to 800 V maximum for conventional Ond-Classic silicon inverters operating, for the same power switched, at a switching frequency of 16 kHz. This DC voltage increase improves the dynamics of the active filter and therefore the filtering quality. This improvement in the dynamics of the active filter is accentuated thanks to the rapid switching of the components of the Ond-Cr inverter.

Advantageously, a saturation element 20 is conventionally configured to impose high and low limits on the control signal. These limits are determined by the amplitude of the carrier wave which is related, in turn, to the voltage of the capacitive energy storage element 3.

Advantageously, the input signal of the saturation element 20 comes from the output of an adder 22, the output signal of which is the sum of a control voltage u emitted by a controller 23 and of the voltage V _S of the connection point C. The addition of the voltage V _s in the control loop of the injected current I _inj prevents a strong inrush of reactive current from the inductance L-Ond-Cr of the filter 11.

The calculation unit 25 of the reference currents

The calculation unit 25 of the reference currents comprises an advanced phase locked loop unit 26 (advanced PLL, in English “ Phase Locked Loop” ), which extracts, in addition to the angle provided by a locking unit conventional phase, the amplitude of the direct component of the voltage at the connection point C of the given electrical network 1.

This advanced PLL has the ability to operate in a medium/network already disturbed in harmonics and voltage unbalance, with two center frequencies (50 and 60 Hz 10%) as well as with a variation in fundamental voltage amplitude of 15 %, to cover the operation of the given electrical network 1 and generators.

Advantageously, the advanced PLL is adapted to operate in the electrical network at a fundamental frequency that varies between 360 Hz and 800 Hz, advantageously the fundamental frequency is 400 Hz, for a voltage comprised between 115 and 200V.

Advantageously, the electrical network is an on-board electrical network such as an airplane, a ship, etc.

Advantageously, the fundamental voltage is a phase-to-phase voltage.

Advantageously, the advanced PLL 26 allows:

• to ensure equality between the current absorbed by the capacitive energy storage element 3 and the maximum power current produced by the renewable energy generation power unit 100; this is thanks to the extraction of the amplitude of the direct component of the voltage of the connection point of the given electrical network 1;
• to ensure effective behavior of the method for identifying disturbance currents for a given electrical network voltage 1 already highly distorted in harmonics and unbalanced;
• to ensure current injection of the maximum power of the renewable energy generation power unit 100 in phase with the direct component of the voltage of the connection point C of the given electrical network 1, thereby preventing a additional consumption of reactive power through the given power grid 1.

This allows correct operation of the compensation device 7 even in the case of generating sets replacing the given electrical network 1 or the on-board networks.

Advantageously, the compensation device 7 ensures synchronization between the maximum fundamental current generated by the renewable energy generation power unit 100 and the direct voltage component of the given electrical network 1 of the connection point C. This avoids, d on the one hand, an additional consumption of the reactive power and makes it possible, on the other hand, to ensure a charging current of the capacitive energy storage element 3 not exceeding the maximum current of the power unit at renewable energy generation 100.

The calculation unit 25 of the reference currents comprises a plurality of measured inputs.

The quantities of the system that are measured include at least:

- the intensity I _L of the current of each wire consumed by the load 2; and or

- the intensity I _inj of the current of each wire leaving the filter unit 10; and or

- the intensity I _pv of the current of the renewable energy generation power unit 100; and or

- the voltage V _pv of the renewable energy generation power unit 100; and or

- the voltage V _dc across the terminals of the capacitive energy storage element 3, which is also the voltage of the maximum output power of the renewable energy generation power unit 100; and or

- the voltage V _S of each of the phases at the connection points C, via an advanced phase-locked loop (PLL) 26 which is integrated in the calculation unit 25. The phase-locked loop of advanced phase 26 is used to extract the three-phase positive or direct component V _d123 from the voltage of the given electrical network 1 at the connection points C. Advantageously, the voltage of the given electrical network 1 has a three-phase positive component V _d123 useful for a efficient operation of the computing unit 25.

The advanced PLL offers, in addition to the phase provided by a classic/conventional PLL, the amplitude (V _d ) of the direct component of the connection voltage of the given electrical network 1, as described in Figures 4 and 25. direct voltage ensures effective performance of the method for identifying disturbance currents, even in the case of operation of the active filter in a disturbed environment of the given electrical network 1 (distorted network voltage with the presence of harmonics) and/or unbalanced . In addition, when regulating the voltage of the DC bus of the inverter, extracting the amplitude of the voltage direct component of the given electrical network 1 allows the charging of the capacitor by an equal optimum current, in the case of a renewable energy generation, aware of the maximum power of the renewable energy generation power unit 100. Advantageously, this current of the maximum power Imp will be injected into the given electrical network 1, thanks to the advanced PLL, in phase with the positive sequence component of connection point C. This prevents additional consumption of reactive power.

In a possible aspect of the invention, the calculation unit is split into three:

- a reference current calculation unit 25-A (Upstream) configured to determine the non-active disturbing current flowing in load 2, and

- a reference current calculation unit 25-B configured to calculate, based solely on the power, the voltage of the maximum active power point of the renewable energy generation power unit 100,

- a reference calculation unit 25-A (Downstream) delivering the non-active/disturbing currents of unit 25-A (Upstream) and/or the maximum active current from unit 25-B and regulator 62, which ensures at its output the maximum power of the renewable energy generation power unit 100.

The current to be injected at connection point C comprises the active current of the renewable energy generation power unit 100 as well as disturbing non-active currents, which may present harmonics, unbalance, and reactive, configured in opposition to phase to oppose the harmonics, unbalance and reactive of the signal flowing through the load 2 so as to reduce or even cancel them on the given mains side 1.

The switching control of the inverter 9 Ond-Cr is carried out in such a way as to let pass through the output filtering unit 10 at the connection point C at least:

• some or all of the non-active disturbing currents injected in phase opposition comprising: harmonic currents, as well as reactive and unbalanced currents at the fundamental frequency, in the non-linear and linear electrical loads 2,

to satisfy the demand for non-active energy consumption of non-linear and linear electrical loads 2, while depolluting the given electrical network 1 from these non-active disturbance currents; and or

• an active current, in phase with the direct or positive component of the voltage of the connection point C thanks to the advanced PLL 26, corresponding to a point of maximum power available within the power unit with renewable energy generation 100 , to satisfy the active power consumption demand of non-linear and linear electric loads.

The controller 23 ensures the continuation of the current between the injected current I _inj and the reference current from the unit 25. For its part, the regulator 62 ensures the continuation of the voltage between the voltage V _dc at the terminals of the element of capacitive energy storage 3 and the voltage of the maximum power of the renewable energy generation power unit 100, delivered by the reference current calculation unit 25-B; the capacitive energy storage element 3 will therefore be charged.

Finally, the switching control device 21 controls the switches of the inverter 9 in order to generate I _inj .

Advantageously, the reference current calculation unit 25 comprises an output for each phase, each output corresponding to the control-command chain of the associated phase (including the neutral), said control-command chain comprising a comparator 24 , a controller 23, an adder 22, a saturation element 20 and a logic comparator 18 whose output is divided into two branches, one of which comprises a logic level inverter 19.

The calculation unit 25 of the reference currents therefore emits a setpoint signal per phase, including the neutral, having an intensity I _ref .

The intensity I _inj of each phase, including the neutral, of the signal emitted by the inverter 9 is returned to the control-command unit 12 and compared, via the comparator 24 with the setpoint I _ref of the control chain command of the corresponding phase, delivered by the calculation unit 25. The difference between I _ref and I _inj is corrected via the controller 23 which issues the command u.

Advantageously, the voltage V _dc applied to the terminals of the capacitive energy storage unit 3 is regulated (maintained constant) by comparing it, via a comparator 60, with a reference voltage V _dc-ref , which is equal to the voltage of the maximum power point V _MPP of the renewable energy generation power unit 100, calculated by the calculation unit 25-B, described as in the .

In a possible aspect of the invention, the measured signal of the voltage V _dc is filtered from fluctuations at 300 Hz or other fluctuations, via a second order low-pass filter 61. The error signal (V _{dc- ref} - V _dc ) is controlled by a PI (Proportional Integral) regulator 62 or another suitable controller, in order to obtain the maximum power P _MPP .

The voltage V _S of the given electrical network 1 is added in the adder 22 to the command u, the output of the adder 22 being limited in the saturation element 20, the output of the saturation element 20 being compared via the comparator 18 logic with the signal delivered by the pulse width modulation device (the carrier).

The output signal of logic comparator 18 is at a level 1 if the output signal of saturation element 20 is greater than the carrier. Otherwise it is at level 0. According to this logic chain, the switching devices 13 or 14, which do not include a logic level inverter 19, are respectively closed or open (the other device operating in a complementary manner) .

The control-command unit 12 comprises at least one processor and at least one memory, the memory comprising a program executed by the processor so as to implement the method for determining the setpoint signal I _ref , containing the non-active current as well as the current I _MPP of the maximum power point MPPT, via the calculation unit 25 of the reference currents, of the control of injected current I _inj via the controller 23, to regulate the voltage V _dc at the terminals of the unit of capacitive energy storage 3 via the regulator 62 and to generate this injected current by controlling the switching devices 13, 14 of each wire/arm of the inverter 9 via the switching control device 21.

According to one of the aspects of the invention, the processor required, thanks to the simplified controller 23 associated with the L-Ond-Cr output filter 11 and the Ond-Cr inverter 9, is much less powerful than that required by the nonlinear control. quite sophisticated associated with an LCL type output filter and an inverter with classic Ond-Classic silicon components.

The processor is a digital implementation target.

The method for determining the reference signal I _ref , delivered by the calculation unit 25 of the reference currents, comprises the following steps:

- measurement processing (based on the instantaneous or other 25-A power identification algorithm) of the current I _L of load 2 and of the voltage at the connection point V _s (including the direct voltage component (in amplitude V _d and in phase θ _d ) is extracted by the advanced PLL) so as to estimate the non-active harmonic, reactive and unbalanced currents;

- generation of a setpoint I _ref calculated to cancel the non-active content of the current I _s on the given electrical network side 1.

Advantageously, the reference current calculation unit 25 is configured to identify, via the unit 25-B, the point of maximum operating power MPPT of the renewable energy generation power unit 100, installed within of a renewable energy production field, a building or a self-consumption plant 99, as follows:

- processing of the measurements of the current I _pv and the voltage V _pv (based, via unit 25-B, on a P&O type algorithm from English " Perturb and Observe" or others, which is part of unit 25 ) of the renewable energy generation power unit 100.

Advantageously, the P&O algorithm of the calculation unit 25-B is adapted to an embodiment of the invention, which ensures connection without a chopper of the renewable energy generation power unit 100 to the element capacitive energy storage 3; the output of the P&O algorithm is, in this case, the maximum power voltage V _mpp and not the chopper control signal, usually associated with the maximum power detection algorithm, P&O or others.

- generation of the maximum voltage V _MPP , via the unit 25-B, and consequently the maximum power P _MPP and the maximum current I _MPP which will be integrated into the setpoint current I _ref via the power P _MPP calculated at the output regulator 62.

Advantageously, only the algorithm for calculating the maximum power point (unit 25-B) is to be integrated into unit 25, via the algorithm for identifying disturbance currents (unit 25-A). Indeed, the capacitor voltage regulation loop (capacitive energy storage element 3) of the inverter is already provided in the 25-A unit to charge the inverter capacitor during pure filter operation active, in order to compensate for the losses caused by the components of the Ond-Cr inverter as well as the L-Ond-Cr filter.

According to one of the aspects of the invention, during operation as an active filter with injection of the maximum power from a renewable energy source, the reference voltage Vdc-ref becomes the voltage of the maximum power, delivered by the unit 25-B, instead of being predefined according to specifications oriented pure filtering. This methodology lightens, minimizes and improves the precision, in a notable way, during the implementation of the control-command part, in comparison with the state of the art which provides for an additional loop of each injected current which imposes also the use of a chopper.

The control-command unit 12 is further configured to perform the regulation of the voltage V _dc across the terminals of the capacitive energy storage element 3, and therefore ensure the charging of the capacitor (storage element d capacitive energy 3). Indeed, the capacitor (capacitive energy storage element 3) has, among other things, the role of covering the losses of the inverter 9 and the filtering unit 10 as well as providing the maximum active current I _MPP of the renewable energy generation power unit 100 to the connection points C, via the Ond-Cr inverter 9 and the filter unit 10 of the L-Ond-Cr filter 11.

Advantageously, the rest of the process within the control-command unit 12 comprises:

- a processing of the measurements of the voltage V _dc at the terminals of the capacitive energy storage element 3;

- filtering, via a second-order low-pass filter 61, of the measured signal of the voltage V _dc , fluctuations at 300 Hz or others;

- the continuation, via the regulator 62, between V _dc-ref =V _MPP and the voltage V _dc , in order to ensure a constant maximum voltage V _MPP across the terminals of the capacitive energy storage element 3;

- the generation of the disturbing current injected in phase opposition with respect to that circulating in the load and of the current of the maximum power of the renewable energy generation power unit 100.

Advantageously, for this tracking via the regulator 62, the calculation unit 25 of the reference currents is configured to integrate the calculation of the current of the maximum power, via the unit 25-B, in the algorithm of the calculation of the currents not assets, provided by unit 25-A.

The process of this integration can follow the following way: the output of the regulator 62 being the maximum power P _MPP of the renewable energy generation power unit 100, this signal is added to the adder 63, which has, at its second input, the active disturbance power from unit 25-A (Upstream) of the calculation of instantaneous disruptive powers (active , reactive and homopolar . Unit 25-A (Upstream) calculates, from the voltages V _s at the connection points C and the currents I _L of load 2, the instantaneous disruptive powers (active , reactive and homopolar in the reference α, β and 0) caused by the non-active disturbing currents present in the current of the load 2 I _L .

Advantageously, the calculation of the setpoint/reference currents I _ref is done via an inverse passage, with respect to the unit 25-A (Upstream), through the unit (25-A Downstream). This reference current contains the non-active currents as well as the maximum power current, first calculated in the same α, β and 0 marker then in the three-phase 4-wire marker.

It should be noted that the advanced PLL 26, used in the disturbance current calculation unit 25-A, extracts both the amplitude V _d as well as the angle θ _d and consequently the three-phase direct component V _d123 of the connection voltage (point C) V _s . Thus, precise identification of disturbing currents even in the case of a network disturbed by harmonics and voltage unbalance is ensured.

Moreover, a simple division of the instantaneous powers, as described in the , by the square module of the voltage V _d123 , calculated in the reference (α,β,0), ensures the injection of the current of the maximum power of the renewable energy generation power unit 100, at the points connection C, in phase with V _d123 of the network voltage V _s . This limits additional consumption of the reactive power via the given electrical network 1. In addition, the direct component amplitude extracted by the advanced PLL 26 ensures equality between the current absorbed by the capacitive storage element 3 and the current of the maximum power produced by the renewable energy generation power unit 100.

The 25-A(Upstream) units including the advanced PLL 26 and 25-A(Downstream) are configured to identify disturbance currents; the units 25-A(Upstream), 25-A(Downstream) and 25-B represent the reference current calculation unit 25.

It should be noted that thanks to the control-command unit 12 and especially the calculation unit 25-A & 25-B with the advanced PLL as well as the controllers 23, the device 7 is adapted to two fundamental frequencies between 40 and 70 Hz, to understand the case of a generator. In addition, the device is suitable for a nominal voltage of the given electrical network 1 between 180 and 480 V.

Advantageously, device 7 is adapted to operate in the electrical network at a fundamental frequency that varies between 360 Hz and 800 Hz, advantageously the fundamental frequency is 400 Hz, for a voltage comprised between 115 and 200V.

Advantageously, the electrical network is an on-board electrical network such as an airplane, a ship, etc.

Then, the control-command unit 12 having the setpoint current I _ref , coming from the calculation unit 25, as well as that injected I _inj controlled in turn via the controller 23, the Ond-Cr inverter 9 is controlled via the switching control device 21 to generate the current I _inj .

Advantageously, the Ond-Cr inverter imposes the voltage on the DC side in this case. Cala represents an additional reliability of the parallel active filter device as well as a significant financial gain.

Indeed, this makes it possible to avoid the use of an additional device, a chopper inserted between the renewable energy generation power unit 100 and the inverter 9 Ond-Cr, for, among other things, generating the power maximum from renewable energy generation power unit 100.

Advantageously, the control-command unit 12 can manage the control of a chopper configured to maintain a predetermined DC voltage across the terminals of the capacitive energy storage element 3 of the Ond-Cr inverter. For example, the control-command unit 12 can comprise:

- a control of a chopper, configured to maintain a constant predefined direct voltage at the terminals of the capacitive energy storage element 3 of the Ond-Cr inverter, independently of the voltage level of the generation power unit of renewable energy 100 to ensure an unchanged harmonic filtering quality;

- a double chopper, for the case of an islanded electrical network. The second chopper can provide, for the case of the island network, micro-grid and for other applications, an additional level of the DC voltage, including for the case of a charge/discharge management of a battery bank or other capacitive energy storage system. The inverter, in the case of an island network, imposes the voltage and frequency on the AC side.

Filter unit 10

The filtering unit 10 comprises in one embodiment a filter 11 of the inductance type with a low inductance L-Ond-Cr (equal to or less than L _f1 +L _f2 of the LCL filter and ten to twenty times less than that of the filter L -Classic both associated with a classic silicon inverter Ond-Classic: around 200 to 500 μH or less of Ond-Cr against 2 to 5 mH of L-Classic depending on the switching frequency).

The output filter retains the filtering efficiency of the harmonic components, due to chopping, of an LCL filter connected to the output of an inverter made of conventional silicon components which generates these harmonic components, the filter L avoiding the resonance between the filter LCL and the given power grid 1.

The filter 11 therefore comprises, on each of the 4 wires (three phases with the neutral), an inductance 15 connected on the one hand to an input wire of the filtering unit 10 (outgoing from one of the 4 connection points common A), and on the other hand to one of the 4 connection points C via an output wire the filter unit 10.

The filter 11 makes it possible to prevent the propagation of the components due to the switching frequency of the inverter 9 to the given electrical network 1 without degrading the dynamics of the compensation system 7. It therefore makes it possible to limit the risk of resonance in the case of an LCL filter as well as in the case of an auxiliary passive filter associated with an L-Classic intended to limit this propagation.

In one possible aspect of the invention, the output filter L is an output filter of the LCL type (with two inductors 15 and 16 connected to a capacitor 17) whose capacitor C is short-circuited, thus Lf1+Lf2=L -Ond-Cr, Lf1 being the inductance of the first coil of the LCL filter and Lf2 being the inductance of the second coil of the LCL filter.

The signal emitted at the output of the filtering unit 10, and therefore at the output of the compensation device 7, therefore presents the harmonics of the signal caused by the non-linear part of the load 2.

For any harmonic present in the disturbing signal, caused by load 2, the signal emitted by the filtering unit 10 presents an inverse voltage value (in phase opposition) to the voltage value of the signal from non-linear load 2 at this harmonic of rank n.

Thus, by injecting the output signal of the compensation device 7 at the connection point C, the harmonics of the signal emitted by the non-linear load 2 are canceled and the current flowing on the given electrical network side 1 is devoid of harmonics.

Similarly, the output signal of the compensation device 7 has, for the fundamental frequency, a phase shift configured to generate an inverse reactive power (in phase opposition) to the reactive power of the disturbing signal of the load 2.

In this way, the reactive power consumed by the load 2 is completely compensated, within the limit of the dimensioning of the device 7, on the given electrical network side 1.

It should be noted that any disturbing current at a frequency included in the frequency band of the injected current control loop, ranging from 0 Hz to 2500 Hz, can be compensated/filtered by device 7, as described in . Unconventional harmonics can be included.

The Capacitive Energy Storage Element 3

The capacitive energy storage element 3 comprises the capacitor or capacitors connected upstream to the DC input of the inverter and downstream to the DC output of the renewable energy generation power unit 100. The voltage DC applied to the terminals of the capacitive energy storage element 3 V _dc is to be kept constant by the control-command unit 12. This voltage may be greater than 1000 V when using the Ond- Fast component Cr (above 80 kHz), which favors the overall dynamics of the active filter.

It should be noted that for the same sizing power as Ond-Cr, limitations of an industrial nature are imposed on the classic silicon inverter, in terms of switching frequency at 16 kHz and DC voltage at 800 V.

A renewable energy generating device ensures the production of the maximum active power of the renewable energy generating power unit 100, while offering the possibility of compensating all or part of the reactive power and/or the unbalanced current. This operating mode provides, as an option, charge/discharge management of the energy storage system (batteries or other energy storage technologies).

In this embodiment, the device operates as a renewable energy system. The device in this case injects only the maximum power of the renewable energy generation power unit 100, with the possibility of compensating reactive power and current imbalance.

In this embodiment, the output filter 11 can be an LCL or L-Ond-Cr with L-Ond-Cr less than or equal to L _f1 +L _f2 of the LCL filter.

In this embodiment, active filter operation may be an option.

The renewable energy production device can be associated with the parallel active filter, the management of the charge/discharge of the energy storage system (batteries or other energy storage technologies) is an option.

In another aspect of the invention, there is proposed a current compensation device 7 of the parallel active filter type, able to be connected:

• at its input, downstream of at least one renewable energy generation power unit 100 coupled to a capacitive energy storage element 3, and
• in parallel, at its output, upstream to a connection point C between on the one hand a given electrical network 1, and on the other hand non-linear and linear electrical loads 2,

the current compensation device 7 having:

• a power conversion unit 8, comprising at least one fast-switching voltage structure power inverter 9 with silicon carbide SiC or gallium nitride GaN components,

the conversion unit 8 generating an alternating current at the frequency of 50 Hz, covering the maximum active current generated by the renewable energy generation power unit 100;

• an output filtering unit 10, comprising an output filter 11 for each of the phases and a neutral, and connected: on the one hand downstream of the inverter 9, and on the other hand in parallel to the connection point C between the given electrical network 1 and the non-linear and linear electrical loads 2, the output filter 11 being sized to block the harmonic components due to the rapid switching of the inverter 9;

the output filter is an L-type inductance first-order filter, avoiding resonance between an LCL filter and the given power grid 1 and at least 10 times lower than the inductance of a first-order filter connected at the output to a silicon component inverter;

• a control-command unit 12 comprising a unit 25 for calculating reference currents, the reference currents comprising:

• at least one active current corresponding to a maximum power point of the renewable energy generation power unit 100;
• at least one active current for recharging the capacitive energy storage element 3,
• advantageously: at least one current generated by the non-linear and linear loads 2,

presenting all or part of the reactive current and/or the current unbalance at the fundamental frequency,

the control-command unit 12 also comprising:

- a 1st order continuous sliding mode controller with an approximated sign function in Sigmoid function; or an improved linear controller of the RST _am type.

- a switching control device 21 which controls the switching of the inverter 9 and which provides closed-loop control at the fundamental frequency 50 Hz for the injection by the inverter of the non-active disturbance current (reactive and unbalanced) and the active current, depending on the identification of the reference currents by the calculation unit 25,

- a PI type regulator 62, which ensures at its output the maximum power of the renewable energy generation power unit 100,

the switching control of the inverter 9 being carried out in such a way as to let pass through the output filter unit 10 at the connection point C:

• an active current corresponding to a maximum power point available within the renewable energy generation power unit 100, to satisfy the active energy consumption demand of the non-linear and linear electrical loads 2, while ensuring the charging the capacitive energy storage element 3.
• advantageously: some or all of the non-active disturbing currents injected in phase opposition comprising reactive and unbalanced currents at the fundamental frequency, in the non-linear and linear electrical loads 2,

to satisfy the demand for reactive energy consumption of non-linear and linear electrical loads 2, while decontaminating the given electrical network 1 of these non-active disturbance currents.

Advantageously, the device has only two voltage levels.

In another aspect of the invention, one of the embodiments of the compensation device 7 is included in an electrical system comprising a given electrical network 1, non-linear or linear loads 2.

Advantageously, the system is connected upstream to a renewable energy generation power unit 100.

The renewable energy generation power unit 100 can be:

• chosen from the following list: one or more photovoltaic panels, wind turbine(s), fuel cell(s), and
• coupled, directly or via a DC/DC power electronics device to a capacitive storage element 3 in the case of continuous production, or via an AC/DC power rectifier in the case of AC production.

The given electrical network 1 can be chosen from the following list: the main electrical network, a local electrical micro-network islanded or connected to the main electrical network, or an on-board electrical network.

In one embodiment, the system further includes a smart building 27.

Advantageously, one of the embodiments of the compensation device 7 is installed within the so-called smart building 27, that is to say that the electrical appliances contained in the smart building 27 can, among other things, be controlled and activated selectively by a decentralized management unit 70, for example to operate during so-called off-peak periods of the day, during which the energy demand of the given electrical network 1 is low and the cost of energy, from the consumer's point of view, decreases .

The control-command unit 12 is connected to the decentralized management unit 70 of the intelligent building 27. The decentralized management unit 70 communicates in real time to the control-command unit 12 the powers of the loads (electrical appliances) in working order or not of building 27; the control-command unit 12 can compare the maximum available power of the renewable energy generation power unit 100 with the total load of the intelligent building 27.

The control-command unit 12, having in real time the maximum power of the renewable energy generation power unit 100 delivered by the unit 25-B and the regulator 62 as well as the powers of the loads in the state of operation or not of the building 27 communicated by the decentralized management unit 70, is configured to regulate the consumption of the various devices according to at least two modes of economic distribution of the loads.

Advantageously, the control-command unit 12 is configured to optimize the consumption of the various devices operating within this intelligent building 27 by distributing the loads corresponding to the non-linear/linear loads according to at least two operating modes:

• a first distribution mode, called adapted consumption mode, in which the control-command unit 12 controls the decentralized management unit 70 of the intelligent building 27 so as to adapt the consumption of the intelligent building with the production of the unit of power at renewable energy generation 100, so that the total load curve of the smart building has a maximum simultaneity factor corresponding to the operation of all the payloads of the building at the same time, within the limit of the renewable energy produced,

the first mode has charge/discharge management of the batteries or other energy storage systems within the same positive energy smart building 27 or between the buildings 27 interconnected via the control-command units 12 of the compensation device 7 ( this first mode of operation provides, according to the strategy adopted by the producer/cons-producer);

• in the event of insufficient renewable energy production, a second distribution mode, called modulated consumption mode, in which the control-command unit 12 controls the decentralized management unit 70 of the intelligent building 27 so as to modulate the consumption of the devices of the intelligent building 27 to tend to a substantially constant total load curve of the intelligent building as a function of time;
• in the same case and if the given electrical network 1 is intelligent, a third mode of distribution is adapted to the production of the electrical network 1 is possible.

We denote by the payloads of the building, the maximum quantity of electricity supported by this given electrical network 1.

Advantageously, the system further comprises:

• a local network 28 connected to the given electrical network 1, and
• 80 conventional power generation units, and
• a semi-decentralized management system 29, and
• a plurality of power units with renewable energy generation 100 connected to the local network 28, by a compensation device 7, each compensation device 7 being connected to the semi-decentralized management system 29 to which it communicates information concerning the production of energy, present and future, of each of the 100 renewable energy generating power units, and
• a plurality of consumption stations corresponding to the non-linear and linear loads 2, each of the consumption stations being connected to the local network 28 and equipped with a compensation device 7, connected to the semi-decentralized management system 29 to which it communicates information concerning instantaneous consumption and future consumption according to the programmed operation of the consumption stations,
• a plurality of power units with renewable energy generation 100 and linear and non-linear loads 2 of the smart buildings 27 with positive energy, which ensure self-consumption and where any surplus energy is stored or exchanged with the others intelligent buildings or delivered to the local network 28 via the control-command unit 12 of the compensation device 7, in coordination with the semi-decentralized management system 29 in the event of an exchange with the local network 28.

Conventional energy production units refer to nuclear energy and fossil fuels such as coal, gas, oil, etc.

In one embodiment, the semi-decentralized management system 29 is configured to ensure the economic distribution of production from the conventional energy production units 80.

According to the present invention, the terms “economic dispatch”, “economic dispatching”, “ Economic Dispatch” and “ Unit Commitment” are used interchangeably by those skilled in the art.

Advantageously, the semi-decentralized management system 29 is configured to intervene when the total demand for energy within the local network 28 is greater than the total production, with the decentralized control units 70 of the intelligent buildings 27, via the control units control-command 12 of the compensation devices 7, to switch them to a modulated consumption mode.

Advantageously, the semi-decentralized management system 29 is configured to control the distribution on the local network 28 of the power coming from the given electrical network 1 if the estimated total production does not cover the demand.

In a first mode of economical load distribution, the intelligent building 27 comprising a renewable energy generation power unit 100 and in the event of high availability of renewable primary sources, the control-command unit 12 is configured to control the decentralized management unit 70 so as to activate most of the useful devices of the intelligent building 27 according to the production of the renewable energy generation power unit 100; this mode of consumption is called adapted mode.

This mode of distribution provides for the storage of excess energy in a battery bank or in another energy storage system.

In a second embodiment, where the renewable primary energy is not very available, the control-command unit 12 is configured to control the decentralized management unit 70 so as to selectively activate the devices of the intelligent building 27, so as to to present a flat or other load curve, depending on the producible power of the given electrical network 1; this mode of consumption is called modulated mode.

This makes it possible to avoid peaks in demand on the given electrical network 1 and to benefit from a reduced cost of the energy consumed.

In this second embodiment, management of the energy stored in the batteries (or in other energy storage systems) is provided to supplement the energy of renewable origin produced by the generation power unit of renewable energy 100 and best meet the demand of the activated loads of building 27.

In one embodiment, the compensation devices 7 as well as the decentralized management units 70 of the neighboring buildings 27 are interconnected, which makes it possible to ensure management of the charge/discharge cycles of the batteries or other energy storage systems. energy installed in buildings 27.

In a first mode of operation, the surplus energy of the renewable energy generation power unit 100 installed within a building 27 can be supplied, via the units 12 of the devices 7 and the management units decentralized 70 concerned, to other neighboring buildings 27, to cover their energy demands and/or to store energy in their storage systems (batteries or other).

In a second mode of operation, the energy stored in a storage system (batteries or other) of a building 27 can be supplied, via the units 12 of the devices 7 and the decentralized management units 70 concerned, to the other buildings 27 neighbours, to cover their energy demands and/or to store energy in their storage systems.

In a third mode of operation, the energy stored in the storage systems (batteries or other) of the buildings 27 can be supplied, via the units 12 of the devices 7 and the decentralized management units 70 concerned, to the local network 28 via the semi-decentralized management system 29 to which the compensation devices 7 are computer-connected.

Most or even all of the energy consumed is, in this case, of renewable origin, which makes it possible to minimize the consumption of energy from the given electrical network 1. In the event of a strong need for power supply and in complement of the second distribution mode and if the given electrical network 1 is intelligent, a fourth distribution mode adapted to the production of the given electrical network 1 intelligent is possible.

In one embodiment, the compensation device 7 is installed between the given electrical network 1 and an industrial site or a residential, administrative or commercial building, each of these frames is comparable from the point of view of the given electrical network 1 to a load nonlinear/linear disturbance 2.

In one embodiment, the system comprises a plurality of renewable energy generation power units 100, for example of the wind or photovoltaic type, each connected, directly or via a device 7, to a local network 28, itself even connected to the given power grid 1.

The local network 28 corresponds to producers, consumers and consumer-producers (self-consumption or positive energy buildings).

Most of the time, each renewable energy generating power unit 100 is connected to the local network 28 by means of a compensation device 7 configured to maximize the power production of the renewable energy generating power unit. renewable 100 as well as to prevent the propagation of electrical disturbances, in the event of their presence, on the upstream side of the compensation device 7 towards the local network 28 and the given electrical network 1.

Each compensation device 7 is connected, by computer, to a semi-decentralized management system 29, to which it communicates, in real time, information concerning the consumption as well as the energy production of the power units with renewable energy generation. 100 actual and forecast.

In addition, the semi-decentralized management system 29 receives real-time information on the energy producible from the conventional (fossil) energy production units 80, installed within the local network 28.

The semi-decentralized management system 29 manages the economic distribution of conventional (fossil) energy generators within the local network 28, according to the total production of renewable origin as well as the total consumption within the same local network 28.

The semi-decentralized management system 29 receives information in real time on the powers of renewable and conventional origin available at any time on the local network 28.

The system also comprises a plurality of consumption stations, for example an intelligent building or an intelligent industrial site, each consumption station being comparable from the point of view of the network to a non-linear/linear load 2, each of the consumption stations being connected to the local network 28, is equipped with a compensation device 7 configured to clean up the current circulating on the local network 28 side of the disturbances caused by the load 2 and to control the activation of the various devices of the consumption station via the decentralized management 70.

Each compensation device 7 is connected to the semi-decentralized management system 29, to which it communicates information concerning the instantaneous consumption and the consumption to come, according to the programmed operation of the consumption stations, so as to estimate the energy demand.

The semi-decentralized management system 29 receiving at any time all the data (actual and forecast) of the energy that can be produced as well as that to be consumed by the various actors of the local network 28 (producer, consumers, consumer-producers), it can intervene, only if necessary when the total demand for energy within the local network 28 is much greater than the total production, with the pilots of the decentralized control units of the intelligent buildings 27 to switch the consumption mode to modulated mode (flat load curve) for the benefit of the overall system.

A consumer-producer station/actor is understood to mean a self-consumption or positive energy building. The semi-decentralized management system 29 is therefore configured to estimate, over a given period, the total energy demand of the local network 28 that it supervises.

Depending on the power production and energy demand estimates, the semi-decentralized management system 29 is configured to:

- carry out, in real time, the economic distribution of production to conventional energy production units 80,

- intervene, only if necessary when the total demand for energy within the local network 28 is much greater than the total production, with the decentralized management units 70, via the control-command units 12 of the devices 7, to make switch the consumption mode to modulated mode (flat load curve),

- control, as a last resort and in the event of a risk of production deficit, the circuit breaker 90 to allow energy consumption on the part of the given electrical network 1. This mode of operation is not possible when the local network is islanded, c i.e. not connected to the given electrical network 1.

The management system 29 is therefore empowered to prioritize the operation of certain consumption items in relation to others, in order to distribute the energy demand over the given period.

In addition, the semi-decentralized management system 29 can spread the power demand over time so that when power from the given electrical network 1 is needed, it is consumed during periods of low demand, so as to minimize costs and avoid loading the given electrical network 1 during demand peaks.

It should be noted that the maximization of the power of the power units with renewable energy generation 100 as well as the optimization of the consumption, within an intelligent building 27, are ensured locally by the devices 7 via the units of control-command 12. The semi-decentralized management system 29 is so called because it is called upon only to ensure the economic distribution of production from the conventional energy production units 80 and to rectify consumption if necessary.

Comparison Invention/Prior Art

1. General structure of the parallel active filter of the prior art

General topology

There presents the general structure of the parallel active filter, which is presented in the form of two blocks: the power circuit and the control circuit. The power circuit consists of:

- a voltage inverter with two voltage levels based on silicon power switches, controllable on starting and blocking (GTO, IGBT, etc.), with antiparallel diodes,

- a capacitive energy storage circuit (a DC bus),

- an output filter.

The control-command circuit is made up of:

- the method for identifying disturbed currents,

- the PLL-based system (Phase Locked Loop ) which is integrated into the method for identifying disturbance currents,

- regulation of the DC voltage applied to the energy storage elements,

- control of the current injected into the given electrical network 1 from the voltage inverter,

- voltage inverter control.

Output filter

The output filter is a passive filter used to connect the voltage inverter to the given electrical network 1. The output filter is sized to satisfy the following two criteria:

• ensure the dynamics of the current

with I _{h_L} the harmonic current contained in the current of the load I _L and I _inj the current of the active filter injected into the given electrical network 1,

• prevent components due to switching from propagating on the given electrical network 1.

Two types of output filter are usually used: a first-order output filter and a third-order output filter, both associated with an inverter with two voltage levels with conventional silicon components (Ond-Classic: IGBT, GTO , etc.) whose switching frequency is of the order of 16 kHz.

First Order Filter (L-Classic)

This type of filter is the most used in the literature; it is composed of a simple inductance L _f of practically negligible internal resistance.

A filter of this type does not make it possible to simultaneously satisfy the two dimensioning criteria of the output filter. Indeed, only a relatively low value of L _f can achieve good dynamics of the active filter by satisfying the equality above.

Unfortunately, a low value of L _f allows the majority of the components due to switching to be found on the given electrical network side 1 and consequently to affect the electrical installations and equipment.

Conversely, a relatively high value of L _f will prevent these components from propagating on the given electrical network 1, but will affect the dynamic range of the active filter and will then degrade the quality of compensation.

The correct sizing of the first-order output filter will therefore depend on the compromise to be found between the dynamics and the efficiency of the parallel active filter.

In order to achieve this compromise, the L-Classic inductor is heavy, bulky, expensive and difficult to size. Indeed, the Ond-Classic & L-Classic active filter on the market, sized to inject a current of 50 A, has a weight of 75 kg, mainly caused by the output filter.

In order to alleviate these constraints and overcome this very difficult to fix compromise, the use of an auxiliary passive filter installed at the output of the inverter or upstream of the given electrical network side 1 is proposed.

However, this auxiliary filter can cause unwanted side effects like resonance with other passive elements installed on the given power grid 1.

It also causes active power consumption through its snubber resistor.

In addition, the filtering quality of these auxiliary filters deteriorates over time due to the aging of their passive elements.

Third-order filter (LCL): modeling in the s-plane

The third-order output filter is an alternative to a heavy, bulky, expensive and difficult to size first-order (L-Classic) filter, allowing to escape the problems mentioned in the case of the first-order output filter.

This output filter is made up of two inductors (L _f1 , L _f2 ) with respective internal resistances (R _f1 , R _f2 ) and a capacitor C _f with a small damping resistor R _f (see ) which we will neglect in the following.

It should be noted that (L _s , R _s and e _s ) represent respectively the inductance, the resistance as well as the electromotive force of the given electrical network 1 upstream.

This type of filter and thanks to the additional degree of freedom, provided by the capacitance C _f , for inductances L _f1 , L _f2 economical, light and not bulky: the sum L _f1 +L _f2 is ten to twenty times less than L- Classic, can ensure the two criteria for sizing the output filter mentioned above.

The equations that model the output filter are:

(Math. 1)

with V _f the output voltage of the inverter, B ₁ (s)/A(s) the transfer function of the output filter with the given electrical network 1 corresponding to the original system (to be controlled) and B ₂ ( s)/A(s) the transfer function corresponding to the disturbance model.

These disturbances are caused by the voltage of the given electrical network 1 e(s) which is now considered equal to the connection voltage Vs for electrical networks with high short-circuit power.

with

From the previous relations, if we neglect all the resistances (except R _f ), we can establish the following relation, valid at frequencies above 50 Hz:

The resonant frequency fcp of the LCL filter, if we neglect in this case also the resistance R _f , is given by the relation:

The LCL output filter is sized to reject the components due to the switching frequency of the inverter, which has been set at 16 kHz, to correspond to an industrial application case.

Thus, a rejection of more than –50 dB is obtained for a cutoff frequency of 1900 Hz.

This choice makes it possible to attenuate the high-frequency components well, as shown by the diagram of the gain of the output filter as a function of the frequency of the .

Nevertheless, the LCL type filter involves a very sophisticated and quite complex nonlinear control to manage as well as, consequently, a fairly powerful calculation unit. In addition, the control of antiresonance between the LCL filter and the given electrical network 1, despite the solution proposed in the patent (WO 2020/007884 A1), remains a point of which manufacturers are skeptical.

1. General structure of the parallel active filter according to the invention

Fast switching component inverter (Ond-Cr)

This inverter is based on fast-switching power electronic components of the order of 80 to 100 kHz or more, depending on the evolution of technology. These components are silicon carbide (SiC) transistors and diodes, the most common currently being SiC MOSFETs. These components can also be of HEMT technology (“High electron mobility transistor”, in French “High Electronic Mobility Transistor”) in gallium nitride (GaN) or other technologies. This Ond-Cr inverter is efficient, reliable, lightweight, less bulky and with good heat dissipation compared to the inverter with conventional silicon components (IGBT, GTO etc.) Ond-Classic.

It should be noted that inverters based on MOSFET components and silicon carbide (SiC) diodes, with continuous or fundamental generation, have already been proposed.

On the other hand, very high frequency inverters based on SiC or GaN or other technologies for harmonic filtering have not yet been studied or produced.

It should be noted that the dynamics of the active filter improves more with the increase of the switching frequency, which promotes the adoption of the Ond-Cr inverter in the structure of the active filter compared to the Ond inverter -Classic. Indeed, the industrial switching frequency of Ond-Cr inverters is 100 kHz, with a rapid and promising increase in power of this frequency, against a high threshold of the switching frequency of active filters, based on an Ond-Classic inverter , marketed from 16 kHz.

First order filter (L-Ond-Cr)

The use of the Ond-Cr inverter allows the use of a first-order output filter of the simple inductance type with minimal constraints (light, economical, not bulky and very easy to size compared to the L-Classic type filter associated with the Ond-Classic inverter) called (L-Ond-Cr). Indeed, a sufficient increase in the switching frequency of the inverter limits the constraints imposed on the output filter, and a simple inductor (with a very low value of the choke: ten to twenty times lower than L-Classic) thus becomes capable of to prevent the propagation of the components of the switching frequency of the inverter towards the given power grid side 1, without affecting the overall dynamics of the system.

In addition, this inductance L-Ond-Cr, whose inductance has a size less than or equal to the sum of the inductances (L _f1 +L _f2 ) of the LCL filter, ensures almost the same retention effect of the components at high switching frequency. than the LCL filter, without risk of resonance and with a much less complex control to implement, for, consequently, a less powerful calculation unit, compared to the LCL filter.

The L-Ond-Cr filter is composed of a simple inductance L _f of practically negligible internal resistance.

According to one of the aspects of the invention, the L-Ond-Cr inductance is sized such that:

L-Ond-Cr L_f1+L_f2=200 to 500 μH, while the L-Classic inductance is between 2 to 5 mH depending on the switching frequency of the Ond-Classic inverter.

Based on the equation (Maths. 1), the models of the system to be controlled (origin system) B ₁ (s)/A(s) and that of disturbance B ₂ (s)/A(s) are given, for the L-Ond-Cr filter as follows. It should be noted that the disturbances are caused by the voltage of the given electrical network 1 which is now considered equal to the connection voltage at C (Vs) for electrical networks with high short-circuit power.

(Math. 2)

with

3. LINEAR AND NON-LINEAR CONTROLS OF THE STATE-OF-THE-ART OND-CLASSIC VOLTAGE INVERTER CONNECTED TO THE GRID VIA L-Classic FILTER

PI Propositional-Integral Controller

The control strategy is based on the estimation of current disturbances by means of an identification algorithm. Then, the Ond-Classic conventional silicon components (IGBT, GTO, etc.) voltage inverter, controlled by the PWM ( Pulse Width Modulation) command, generates the currents injected I _inj to the given electrical network 1, which must follow the identified reference currents I _ref ( ). The closed control loop is designed to ensure high precision tracking.

The given grid connection point voltage 1 Vs here represents an external disturbance, the effects of which are compensated by adding the same given grid voltage 1 to the control signal (u). This will prevent the reactive fundamental current from flowing from the network to the active filter via the L-Classic filter inductor _.

The general diagram of the current control system is shown in the . In this diagram, the Ond-Classic voltage inverter (controlled by the PWM) is connected to the given electrical network 1 via an L-Classic filter, with a controller of the PI type, and the instantaneous or other power method for the identification of current disturbances.

The transfer function of the PI controller is given by the equation:

Based on the formula (Maths.1) with the parameters that models the (L-Classic) output filter (Maths. 2), and using the PI controller, the closed loop system becomes:

with the numerator N(PI)(s) of the transfer function of the PI controller

The parameters K _p and K _i are chosen to ensure good tracking, for the entire band of harmonic frequencies, between the injected currents and those identified, with a minimum response time as well as rejection of disturbances (the network voltage V _s ) maximum.

RST Pole Placement Controller

From equation (1) (Maths. 1) and using an RST controller, we get:

with , And are the polynomials of the controller, as described in the .

The order of And is the same as the order of the L-Classical system ( ); so the polynomials And are first rate. The polynomial is chosen such that for the entire frequency band (50-2500 Hz) included in the reference signal ; can be a simple gain in this case.

It is to highlight that represents the transfer function of the RST controller; the common denominator , named arbitrary stability polynomial, contains the poles of the closed loop. These poles are placed in a sector of the , to ensure a damping of 0.7.

Finally, the poles of the control loop are placed in order to ensure a fast and precise response, with good rejection of disturbances. Note that the pole values are limited by the closed loop cutoff frequency.

Phase shift effect

The RST controller, as well as all the linear controllers, can be used when the references to be tracked consist of constant signals or at a single and relatively low frequency (case of reactive compensation or unbalance at the fundamental frequency 50 Hz). At this frequency, the phase difference between the identified references ( ) and the output of the injected closed loop ( ) is acceptable. On the other hand, if the reference to be followed is composed of signals at several frequencies, the phase shift is no longer negligible. Indeed, the phase shift increases with the frequency. The effect of the phase shift of the structure shown in FIG. 13a is shown in Figure 14. From this figure, we can observe that the disturbed current ( ) is not well compensated for ( : compensation with phase shift), compared to the ideal form ( : compensation without phase shift).

The control loop transfer function with the RST in Figure 13a is presented through the .

The gain and phase of the closed-loop transfer function are given via Table I, for multiples of the fundamental frequency ranging from until . It should be noted that we limited ourselves to 1150 Hz because electrical networks self-filter, via their inductance, high order harmonic currents.

Painting. I. Gain and phase of closed loop transfer function with RST & LCL F(Hz) 50 250 350 550 650 850 950 1150 RST G 1 1 1 1 0.99 0.98 0.98 0.97 pH° -1 -8 -11 -18 -21 -27 -30 -37

Note that the linear controller of RST provides unity gain ( ) for practically the whole frequency band (50-2500 Hz) as described in the . In addition, the RST is only used to compensate for unbalanced and/or reactive currents at the fundamental frequency.

Indeed, at this frequency, a phase shift of (-1°) is negligible. After this frequency, the phase shift is no longer negligible, and the parallel active filter cannot compensate for harmonic currents.

Nonlinear control (by sliding mode - first order)

The Sliding Mode Control (SMC) method is proposed to ensure a desired dynamic response, high robustness / insensitivity to bounded disturbances, good control properties in a wide range of operating conditions as well as very good closed-loop tracking with no phase shift effect.

In practice, the voltage inverter is controlled by a limited/fixed frequency switching function. Thus, variable commutations ( chattering ), if they occur, could lead to overheating which can lead to the destruction of the inverter.

Design of a classic controller by sliding mode for an Ond-Classic and L-Classic

Since the controlled system is first order, then a slip variable is chosen in the form:

(Math. 3)

with

The existence condition of a sliding mode can be easily realized by the SMC

(Math. 4)

with the command applied, which is discontinuous by chattering effect in this case, as described in Figures 16a, 16b and 16c, via the PWM, to the inverter, is fixed and defined by the saturation limiter block at 420V, as described on the .

4- LINEAR CONTROL BY PLACEMENT OF RSTam IMPROVED POLES OF THE OND-CR VOLTAGE INVERTER ACCORDING TO THE INVENTION CONNECTED TO THE GRID VIA THE L-Ond-Cr FILTER

The strategy is based on the estimation of current disturbances by means of an identification algorithm. The change concerns the structure of the inverter, the output filter and therefore the control loop. Indeed, the voltage inverter is with fast switching components (SiC, GaN, etc.) Ond-Cr, controlled by the PWM control, generates the currents injected into the network I _inj through the L-Ond-Cr filter, which must follow the identified reference currents I _ref ( ). The control loop thus closed is designed to ensure high-precision tracking but for an extended bandwidth given the increase in the switching frequency (here 100 kHz).

The network voltage Vs, at the connection point C, here represents an external disturbance, the effects of which can also be compensated by adding the same network voltage to the control signal (u). This will prevent the reactive fundamental current flowing from the network to the active filter via the inductance of the L-Ond-Cr filter _.

The controller used according to one aspect of the invention is the improved RST _am method which proposes a placement of the poles and zeros in order to minimize the error in amplitude but especially in phase between the signal/current identified and that injected.

The improved RST _am controller preserves the same first-order R and S polynomials of the conventional RST, calculated for the L-Ond-Cr system whose model is presented by the equation (Maths. 2), while placing zeros in the control loop in order to minimize, above all, the error in phase.

The Laplace transform of the error is identified:

(Math 5)

with the arbitrary stability polynomial.

Based on the equation (Math. 5), we obtain the following formula:

(Math 6)

To reduce the error to zero, the numerator of eq. (Maths. 6) must tend to zero. Since the polynomials R(s) and S(s) are already defined, only T(s) can be judiciously selected in order to minimize the error . Moreover, the order of T(s) must be chosen so that the transfer function T(s)/R(s) is causal (degree (T) ≤ degree (R)).

(Math. 7)

Using equation (Maths. 7), and replacing s = jω, with ω the angular frequency, the numerator of eq. (Math. 6) becomes:

(Math. 8)

In order to make the right side of eq. (Maths. 8) at zero, the real Re(ω) and imaginary parts Im(ω) must tend to zero. This gives two equations for each frequency.

According to one aspect of the invention, T(s) is a first-order polynomial. Indeed, as the order of R(s) is one, a polynomial T(s) is of first order, which is the maximum order satisfying the causality of the transfer function T(s)/R(s). Given the unknown parameters (t ₁ , t ₂ ) of the polynomial T(s), a single frequency must be chosen. Zero should be placed near the slowest pole of the closed loop transfer function.

In order to show the tracking accuracy of the RSTam controller compared to the classic RST controller, the Bode diagrams, in gain and in phase, of the transfer functions of the control loops ( ) with RSTam and RST superimposed, are plotted on the .

From this figure, we see that the two controllers provide amplitude tracking between the injected current Iinj and the identified one Iref. Moreover, if the phase tracking is well ensured by the RSTam, the RST causes a phase shift between the input and the output of the control loop, significantly degrading the quality of filtering and therefore preventing the applicability of the active filter.

5. NON-LINEAR CONTROL (BY CONTINUOUS SLIDING MODE of first order) OF THE VOLTAGE INVERTER OND-CR ACCORDING TO THE INVENTION CONNECTED TO THE GRID VIA THE L-Ond-Cr FILTER

Since the classic sliding mode controller causes a discontinuous control which results in a chattering effect which can be destructive of the voltage inverter.

Therefore, we employ a classical first-order SMC (with sign function) with continuous control. Indeed, in order to avoid a discontinuous control of the classic SMC, we opt for a sliding mode controller associated with a sign function approximated as a sigmoid function.

Finally, the control of the active filter generated by the DC SMCs will be modulated by a PWM, in order to allow the active filter to operate at a fixed, high switching frequency adapted, on the one hand, to a nominal operation of the fast switching components. (SiC, GaN, etc.: here 100 kHz) power electronics of the Ond-Cr inverter and easy, on the other hand, to be filtered by the L-Ond-Cr, which facilitates, among other things, the blocking of high frequency components by the L-Ond-Cr filter.

In this context, it is important to point out that the Sign function in (Maths. 4) has been approximated as a Sigmoid function with .

6. ADVANCED PLL SYSTEM INTEGRATED IN THE REFERENCE CURRENT CALCULATION UNIT 25 ACCORDING TO THE INVENTION

Based on the three-phase voltage at connection point C ( ), the components V_dand V_qare calculated in the Park representation using the estimated angle of rotation ; with And represent respectively the pulsation and the phase/angle of the forward/positive component of the voltage V_sof the connection point C of the given electrical network 1. This calculation approach is grouped in the following relationship:

Phase locking occurs when . In this case, the angle of the direct voltage component and the estimated one are equal, so we can write, as described in the :

and

It should be noted that the tracking of the phase, including that of the amplitude, is ensured by a controller of the RST pole placement type adapted to the case of an active filter. Indeed, this controller, which must be designed to offer robust performances in a wide band of variation in frequency and voltage (50, 60 Hz 10%), (230/400 15%) respectively, must filter the disturbances external (caused by the unbalance and the harmonic components of the network voltage V _s ) in the closed loop.

In order to achieve these performances, the polynomials R(s) and S(s) are respectively of fifth and sixth order, while T(s) is a simple gain such that .

This special design approach of the adapted RST controller will lead to a change in the methodology of calculating the arbitrary stability polynomial D(s)=A(s)S(s)+B(s)R(s), with the system d origin to be checked, as shown in the :

Indeed, in the polynomial S(s), it is necessary to provide for the elimination of constant disturbances, so s(s)=0. In addition, and in order to eliminate the disturbances linked to the imbalance as well as the voltage harmonics, we include in the polynomial R(s) double zeros at the frequencies 100 Hz and 300 Hz, corresponding respectively to the components due to the imbalance as well as to the first dominant harmonic (the harmonic of rank five). To eliminate the effects of higher order harmonics, a slope of -20dB/decade of the adapted RST regulator has been achieved by choosing a strictly clean RST controller.

Therefore, the polynomial R(s) being of degree 5 and in order to ensure a strictly proper controller Rs(s)/S(s) (degree (R) < degree (S)), the polynomial S(s) is degree 6.

7. SIMULATION RESULTS OF THE CONTROLLED INVERTER ACCORDING TO THE INVENTION

Simulation results with Matlab-Simulink

First-order continuous SMC and RST _am controllers for an Ond-Cr & L-Ond-Cr

The simulations are carried out, initially, via a simple Simulink diagram. In this diagram, the reference harmonic currents identified are modeled by current sources of orders 5, 7, 11, 13, 17, 19, 23 and 25 which represent the same harmonic spectrum of the current that we will analyze via a study of an industrial load below. The simulations are carried out using first the classical SMC (with sign function) then the SMC with a sigmoid function and finally with the improved pole placement method RST _am .

Figure 16c plots single-phase currents , (identified and injected respectively) as well as the inverter control signal . We observe that a very precise pursuit is ensured by the discontinuous control (at variable and very high switching frequency: chattering frequency) of the classic SMC, by the method of SMC with a sigmoid function with continuous control at switching frequency fixed by the PWM as well as by improved linear controller RST _am .

It should be noted that the control by RST _am as well as any linear controller provide continuous control at a frequency adapted to the power electronic components, imposed here by the PWM.

First-order and higher-order SMCs controllers for an Ond-Classic & LCL (State of the art)

The simulations are carried out, for the case of an Ond-Classique associated with an LCL filter, by employing this time, first, in the case of the , the classic SMC then the SMC with sigmoid function then the AIRD method (an artificial increase in the relative degree, followed by an integrator), then in the case of the and in order, the classic SMC, the C-HOSM, the 2-SMC Twisting, the 2-SMC Super-Twisting and finally the Lyapunov approach.

Figures 16a and 16b plot single-phase currents , (identified and injected respectively) as well as the inverter control signal . We can see that very precise tracking is ensured by all the controllers, with a discontinuous command signal for the classic SMC and for the Lyapunov approach.

Parallel Active Filter Simulation Results Installed in a Typical Industrial Network

Two simulation sets will be presented, one corresponds to a pure DPF parallel active filter while the second represents the case of a DPF with the GPV photovoltaic generation option. The industrial non-linear load is a complete six-diode 42 kVA rectifier with a DC-side R//C load and an AC-side inductance of 2 mH.

Finally, the parameters of the given electrical network 1 are quantified in Table II

Table II. Data sheet of the electrical components making up the treated network Electrical Network es, Ssc, Rsc, Lsc 20kV, 500MVA, 0.253Ω, Ls=0.002425H Transformer S _n , ΔP _Cu , ΔPo, u _cc ,I _o 1MVA, 10.5kW, 1.7kW, 6%, 1.3%

The active filter is based, among other things, on the fast switching component inverter using SiC MOSFET components at 100 kHz switching frequency (C3M0021120D), without antiparallel diodes, associated with the L-Ond-Cr=200 μH filter , with the control algorithms, SMC with continuous control associated with a Sign function approximated as a sigmoid function and RST _am .

The technical sheet of the components constituting the active filter is presented via Table III.

Table III. parallel active filter datasheet Parallel active filter L-Ond-Cr output filter L _f =200 µH, R _f = 5 mΩ
Storage capacitor C= 3.7 mF, V _dc =840 V Switching frequency of SiC MOSFET components imposed by PWM 100kHz

It should be noted that the Matlab-Simulink-Simscape Electrical software libraries do not offer models of fast-switching power electronics components. The Matlab-Simulink-Simscape Electrical & PLECS hybrid simulation was used; the PLECS software, which can be installed under MATLAB, offers the possibility of using models of fast-switching power electronic components.

Study of the efficiency of the L-Ond-Cr output filter, comparison with the L-Classic and LCL filters

Figure 17(a, b) shows the spectral analysis of the injected current Iinj in the cases of L Ond-Cr & Ond-Cr and L-Classic & Ond-Classic structures, focusing on the frequency band related to the switching frequency of each case.

From these figures, we find that the efficiency of the L-Ond-Cr inductance associated with the Ond-Cr fast switching inverter, in blocking the components due to the switching frequency exceeds that L-Classic associated with the conventional inverter in Ond-Classic silicon, knowing that L-Classic=2 mH is ten times greater than L Ond-Cr=200 μH. Moreover, the quality of the harmonic filtering is better in the case of L-Ond-Cr compared to the case of L-Classic, for the same controller (THD equal to 1.3% and 2.2% respectively). This superior quality of filtering is due to the faster dynamics of the compensation system provided by the Ond-Cr inverter compared to the Ond-Classic inverter (at respective switching frequencies of 100 kHz and 16 kHz).

Moreover, by comparing the Bode diagrams of the gains of the two output filters LCL and L Ond-Cr, as described in Figures 12 a&b: respectively for the state of the art and the new deposit, we find that the rejection of the components due to the switching frequency (100 kHz) provided by L-Ond-Cr (-42 dB) is quite high/sufficient and approaching the rejection provided by the LCL for an industrial switching frequency of 10 kHz (- 50 dB), knowing that a switching frequency of 16 kHz for conventional Ond-Classic silicon inverters reaches the operating limit of power electronic components. Indeed, at this frequency, accepted by manufacturers, a voltage limitation on the DC side is set at 800 V for conventional inverters, while fast-switching inverters accept a DC voltage greater than 1000 V, which favors the overall dynamic. of the system. In addition, the switching losses at 16 kHz, although they are admissible by the industrialists, reduce the re-demand of the Ond-Classic inverter, while the Ond-Cr inverter is designed for minimum losses for fast switching .

It should be noted that the value of L-Ond-Cr is chosen equal to the sum of the inductances of the LCL: L Ond-Cr=L _f1 +L _f2 =200 μH.

Finally, and by way of example, the shows the case of an undersized output filter (L-Classic=200μH) associated with the classic Ond-Classic silicon inverter with a switching frequency of 16 kHz. From this figure, we see that the output filter, undersized for this switching frequency, is unable to retain the harmonic components due to the switching frequency and most of these components propagate from the inverter to the electrical network. given 1, as depicted in Figures 21a and 21b.

SMC controller continuous to a sigmoid function

In this part, the use of the active filter associated with the Ond-Cr inverter (with SiC MOSFET: C3M0021120D at 100 kHz switching frequency) and the L-Ond-Cr = 200 μH filter with the SMC control algorithm at continuous command associated with a Sign function approximated as a sigmoid function, is validated.

There presents the harmonic spectrum of the load current (the current of the given electrical network 1 without filtering). From this figure, we see that the effective fundamental current of the polluting load is 65 A and the THD of the current is 24.29%.

Figure 19a presents the simulation of phase 1 (before and after filtering) of the current on the network side of Is, of the identified currents Iref and injected currents Iinj as well as the THD of the current ( ).

In this case of simulation, the active filter is initiated after 6 periods of the sector (up to 0.12 s).

It can be deduced from Figure 19a that the harmonic distortion after filtering is 1.3%. This significant reduction in current THD is confirmed by the spectral analysis presented via the .

It should be noted that the THD of the current before the activation of the active filter was 24.29%.

Controller by placement of poles and zeros RST _am

In this part, the previous structure (active filter with inverter of fast switching components (MOSFET SiC: C3M0021120D, at switching frequency 100 kHz) associated with the filter L-Ond-Cr=200 μH), is validated with the algorithm of RST _am improved pole placement control.

Figure 20a presents the phase 1 simulation (with and without filtering) of the mains side current of Is, the identified currents Iref and injected currents Iinj as well as the THD of the current ( ).

In this case of simulation, the active filter is deactivated after 6 periods of the sector (at 0.12 s).

It can be deduced from figure 20a that the harmonic distortion after filtering is 1.77%. This significant reduction in current THD is confirmed by the spectral analysis presented via the .

It should be noted that the THD of the current before the activation of the active filter was 24.29%.

Finally, as an example and in order to understand the importance of this combination of Ond-Cr fast switching inverter and L Ond-Cr output filter, the presents the simulation of the structure of the active filter based on a classic inverter in Ond-Classic silicon with a switching frequency of 16 kHz, associated with an undersized L-Classic output filter of 200 μH. From this figure, we observe the inability of an undersized output filter to prevent the hash components from propagating towards the given electrical network 1. This observation is validated, via the , by a current THD on the grid side given 1 after filtering of 8.20%.

Calculation and integration of the maximum power point in the control-command diagram of the active filter

The objective here is to pursue the maximum power point of a photovoltaic (PV) generator, in order to increase the efficiency of this generation system. There shows the four parameters characterizing the operation of a PV panel/generator; these are the short-circuit current I _cc , the no-load voltage V _co , the maximum power current I _mpp and the maximum power voltage V _mpp (consequently the maximum power available within the generator PV: P _mpp = V _mpp × I _mpp ).

According to one aspect of the invention, the algorithm used for the extraction of the maximum power point is the P&O algorithm (in English “ Perturb and Observe” ), which is based on the disturbance and the observation of the voltage of the PV generator, until reaching the maximum voltage which will correspond to the MPPT point.

This P&O algorithm is adapted to the configuration of the invention which ensures a chopper-free connection of the renewable energy generation power unit 100 to the capacitive energy storage element 3; the output of the P&O algorithm is, in this case, the maximum power voltage Vmpp and not the chopper control signal, usually associated with the maximum power detection algorithm, P&O or others.

DC voltage regulation loop (continuation of maximum power)

We will take advantage of the capacitor voltage regulation loop, on the DC side of the inverter, to ensure the continuation of the maximum power of the PV generator. Indeed, in the case of a parallel active filter which is not connected to a PV generator, the energy storage capacitor regulates itself (charges with a maintenance of a constant voltage) via the given electrical network 1 , through the inverter, to compensate for losses by Joule effect of the power electronics components of the inverter and of the L-Ond-Cr output filter. The voltage of the capacitor V _dc must follow a reference voltage V _dc-ref whose amplitude is chosen to reinforce the dynamics of the system, while respecting the dimensioning of the electrical components of the parallel active filter.

According to one aspect of the invention, the same capacitor voltage regulation loop is used to ensure the tracking of the maximum power point based on the tracking of the power P _MPP and not on the tracking of the current I _MPP . Indeed, unlike the methods of regulating the capacitor of the inverter, which adopt the continuation of the current I _MPP in the case of the presence of a renewable energy generation system, this method will directly extract the maximum power from the generator. PV.

Modeling the Capacitor Voltage Regulation Loop

The relationship between the active power produced by the PV generator and the voltage across the capacitor can be written as:

(Math 30)

The relation (Maths 30) being nonlinear, and for small variations of the voltage V_cdaround its reference V_dc-ref, it can be linearized via the following relation :

(Math 31)

It should be noted that for: (delivered by the P&O algorithm of unit 25-B), we will have . Therefore, the calculated capacitor voltage in the Laplace domain becomes:

(Math 32)

From the relations (Maths 31) and (Maths 32) and by using a linear regulator (Proportional - Integral) or other, the regulation loop of the DC voltage V _dc and consequently the power can be submitted through the . The choice of parameters of this controller ensures fast and precise tracking of the maximum power of the PV generator. It should be noted that the measured signal of the voltage V _dc will be filtered from fluctuations at 300 Hz or others, via a second-order low-pass filter, as described in Fig. .

Integration of the maximum power tracking loop within an active filter

The advantage of the tracking of P _MPP (compared to I _MPP ) is to be able to integrate (with the fewest possible calculation operations) the tracking loop of P _MPP in the algorithm for identifying the disturbance currents of the control-command part of the parallel active filter. In this case, the reference voltage V _dc-ref of the active filter capacitor voltage identification/regulation algorithm is replaced by the maximum power voltage V _MPP (V _dc-ref = V _MPP ) from the algorithm (P & O) adapted as shown in Figure 24. From this figure 24 (a & b), the output of the PI controller being the maximum power P _MPP of the PV generator, this signal is added to the power active disruptive delivered by the 25-A disturbance current identification algorithm; the disturbing instantaneous powers (reactive and homopolar are also calculated. Indeed, this algorithm ensures, initially, the calculation of the disturbing instantaneous powers, caused by the non-active disturbing currents (harmonics, reactive and unbalanced or other) present in the current of the load I _L , in the reference α, β and 0. The calculation of the reference currents I _ref123 is done via a reverse passage, first calculated in the same reference α, β and 0 then in the three-phase reference. These reference currents of the active filter (I _ref123 ) then contain the disturbance currents as well as the maximum current of the PV generator: I _MPP. It should be noted that by basing ourselves on a voltage regulation V _dc whose regulator output (PI for example) is the maximum power, we ensure that the method of instantaneous powers, employed according to one of the aspects of the invention for identify disturbing currents, remains unchanged while guaranteeing the identification of the maximum current of the PV generator: I _MPP in amplitude, with an angle equal to that of the direct voltage component already extracted via the PLL 26 (see Figure 2). The current I _MPP will, in this case, systematically have a phase equal to this angle without passing through the loop of the PLL, because the voltage used in the method for identifying non-active disturbing currents 25-A is that of the direct component . So no special identification method, neither in amplitude nor in phase/angle, dedicated to the extraction of the current I _MPP , as usually adopted in the state of the art.

Indeed, the algorithms which adopt the pursuit of the current I _MPP , instead of P _MPP , would involve the multiplication of I _MPP by three sine functions (shifted by a third of a period) of unit amplitude and angle/phase from the PLL (angle of the direct voltage component of the given electrical network 1 V _s ). This requires more calculations and more precision from the PLL.

According to one of the aspects of the invention, an advanced PLL extracts, in addition to the angle provided by a conventional PLL, the amplitude of the direct component of the voltage at the connection point C of the given electrical network 1.

This the advanced PLL 26, integrated in the unit 25 as shown in the , allow :

• to operate in an environment/network already disturbed by harmonics and voltage unbalance, with two central frequencies (50 and 60 Hz 10%) as well as with a variation in fundamental voltage amplitude of 15%, to cover operation the given electrical network 1 and the generators;
• to operate in an on-board electrical network with a fundamental frequency of 400 Hz, advantageously a variable fundamental frequency between 360 Hz and 800 Hz for a voltage between 115 and 200 V;
• to ensure equality between the current absorbed by the capacitive energy storage element 3 and the maximum power current produced by the renewable energy generation power unit 100; thanks to the extraction of the amplitude of the direct or positive component of the voltage of the connection point of the given electrical network 1;
• to ensure effective behavior of the method for identifying disturbance currents for a given electrical network voltage 1 already highly distorted in harmonics and unbalanced;
• to ensure current injection of the maximum power of the renewable energy generation power unit 100 in phase with the direct component of the voltage of the connection point C of the given electrical network 1, thereby preventing a additional consumption of reactive power through the given power grid 1.

On-board electrical network means an autonomous electrical system having a limited size and having a limited energy consumption, such as on planes, ships, etc.

Result of simulations of the parallel active filter with photovoltaic generation installed in the typical network studied

In this second part, we will validate the use of the active filter with photovoltaic energy generation FAP/GPV. The Ond-Cr inverter as well as the L-Ond-Cr output filter are the same as those used in the first simulation set. Indeed, a structure based on an inverter with SiC MOSFET components at 100 kHz switching frequency (C3M0021120D), without antiparallel diodes, associated with the L-Ond-Cr=200 μH filter, with the continuous SMC control algorithm with sigmoid function are presented.

In this second simulation set, the simulations are carried out using real models of photovoltaic panels (Yingli Solar YL235P-29b module): 30 modules in series and 7 parallel strings which ensure a continuous voltage (Vdc = Vmp) minimum equal at approximately 800 V and maximum equal to 931 V. These voltage values guarantee, with a superiority for the maximum value, a rapid dynamics of the overall system without the need for a chopper (DC/DC converter).

Table IV and the and show the P-V characteristics of the photovoltaic generator (GPV) under STC conditions; the parameters for different irradiations and temperatures are also given by the . Parameters of PV module (Yingli Solar YL235P-29b) Maximum power 213.15W No-load voltage 36.6V short circuit current 7.84A Maximum power voltage 29V Maximum power current 7.35A

This shows that the maximum power, the voltage and the current of the maximum power of the photovoltaic generator GPV, under the STC conditions, where the illumination is 1000 W/m2 and the temperature is 25°, are respectively 44.7615 kW, 870 V and 51.45 A.

Approximate Sign Function to Sigmoid Function Continuous SMC Controller

In this simulation part, we will show only the results of the first-order continuous sliding mode controller with approximated sign function in sigmoid function. Indeed, in the two control cases (RSTam and SMC with sigmoid function), voltage tracking is provided by an independent PI controller and the energy balance depends rather on the algorithm for calculating the maximum power (the P&O adapted ), independent PI controller, non-active current identification method and advanced PLL system.

In this simulation set, the FAP/GPV is tested for harmonic filtering and maximum photovoltaic power generation, with reactive power compensation, under a variable weather profile in both irradiance and temperature. Indeed, two levels of illumination (100 and 1000) W/m2 at 25° C as well as two levels of temperature (10°C, 40°C) at 1000 W/m2, are tested so that the PV generator covers almost all possible weather variations.

There presents the simulation in the time domain of the voltage of the maximum power of the photovoltaic generator Vmp, calculated by the P&O algorithm adapted to the absence of a chopper, as well as the voltage Vdc taken from the terminals of the capacitor on the DC side of the fast switching inverter. The simulation is carried out for the following meteorological conditions: at the beginning, the conditions (100 W/m2, 25° C) are tested for a time belonging to [0, 1] s, then the STC conditions for a time belonging to [1 , 2] s, then the conditions (1000 W/m2, 40 °C) for a time belonging to [2, 3] s and finally the conditions (1000 W/m2, 10° C) for a final slot of temples belonging at [3, 4] s.

It can be observed from the that the adapted P&O algorithm employed, operating in synchronization with the instantaneous power identification unit and the independent DC voltage PI controller, ensures precise monitoring of the maximum power voltage.

Finally, the energy balance is presented via the . In this figure, the reactive powers supplied by the FAP/GPV and the given electrical network 1, as well as those consumed by the non-linear load side, are given.

There validates the active and reactive power balance of the FAP/GPV, even under significant variations in weather conditions. Indeed, when the illumination is very low (for time belonging to [0, 1] s), almost all of the active power is supplied by the given electrical network 1 P _network . Then, for t > 1s and throughout the simulation, the generated active photovoltaic power P _filter-PV exceeds the demand of the load P _load , and the surplus of this power is injected into the given electrical network 1. This results in a positive P _filter-PV and a negative P _network .

Finally, from t> 3s, the FAP/GPV device compensates all of the reactive power of the load Q _load by injecting Q _filter-PV ; the electrical network, which supplied this power to the load, switches to the state Q _network =0 for t> 3s.

The invention offers the advantage of simplifying the method of controlling the current of the inverter.

This invention is the alternative to the two existing solutions.

1. A parallel active filter based on an inverter with two voltage levels, with conventional silicon components (Ond-Classic: IGBT or others), associated with a first-order output filter (an inductor: L-classic) and a easy order. The output filter in this case is heavy, bulky, expensive and very difficult to dimension. Indeed, the output filter, associated with an Ond-Classic inverter, must validate two contradictory criteria, block the components due to the switching frequency without slowing down the overall dynamics of the system. For example, the active filter on the market, sized to inject a current of 50 A, has a weight of 75 kg, mainly caused by the output filter. In addition, this structure limits the voltage on the DC side of the inverter (to 800 V for a switching frequency of 16 kHz and an injected current of 30 A) which limits the dynamics of the filter and consequently the filtering quality.
2. A parallel active filter based on an inverter with two voltage levels, with conventional silicon components (Ond-Classic: IGBT or others), associated with a third-order output filter of the LCL type. The use of this type of filter involves the use of a very sophisticated and quite complex nonlinear command to manage as well as a fairly powerful calculation unit. In addition, the control of antiresonance between the LCL filter and the given electrical network 1, despite the solution proposed in document WO 2020/007884, remains a point of which manufacturers are skeptical. It should be noted that this structure also limits the voltage on the DC side of the inverter (to 800 V for a switching frequency of 16 kHz and an injected current of 30 A) which limits the dynamics of the filter and consequently filtering quality.

One of the aspects of the invention is to propose an L-Ond-Cr output filter (strictly associated with an Ond-Cr fast switching inverter) ensuring almost the same retention efficiency of the components due to the chopping of a filter LCL, without causing anti-resonance or controller complexity issues imposed by LCL. In addition, this inductance is light, not bulky, economical and very easy to size compared to that L-Classic linked to the use of an inverter with conventional silicon components (Ond-Classic). It should be noted that the technology of fast power components allows the increase of the voltage on the DC side of the inverter, compared to the conventional technology in silicon, which benefits the improvement of the dynamics / speed of the system. and therefore the quality of filtering. Indeed, switching frequencies of up to 16 kHz (at a current of 30 A) for conventional silicon inverters Ond-Classique reaches a high operating limit for silicon components. At this frequency, accepted by industrialists, a voltage limitation on the DC side is set at 800 V whereas fast switching inverters accept a DC voltage between 1000 V and 1200 V, which favors the overall dynamics of the system. In addition, the significant increase in the switching frequency of the Ond-Cr inverter (of the order of 100 kHz) also favors the dynamics of the device and consequently the filtering quality, compared to the Ond device -Classic whose maximum frequency is 16 kHz.

Finally, the switching losses at 16 kHz, although admissible, reduce the re-demand of the Ond-Classic inverter, while the Ond-Cr inverter is designed for minimal losses by fast switching.

Claims (32)

Current compensation device (7) of the parallel active filter type, capable of being connected:

• at its input, downstream of at least one capacitive energy storage element (3), and
• in parallel, at its output, upstream to a connection point (C) between, on the one hand, a given electrical network (1), and, on the other hand, non-linear and linear electrical loads (2),

characterized in that the current compensation device (7) of the parallel active filter type has:

• a power conversion unit (8), comprising at least one fast-switching voltage structure power inverter (9) with silicon carbide (SiC) or gallium nitride (GaN) components,

the conversion unit (8) generating an alternating current, with a frequency band ranging from 50 to 2500 Hz, covering the entire frequency band of a non-active disturbing current which presents: all or part of the harmonics, and the fundamental frequency all or part of the reactive power and/or the current imbalance;

• an output filtering unit (10), comprising an output filter (11) for each of the phases and a neutral, and connected: on the one hand downstream of the inverter (9), and on the other hand in parallel at the connection point (C) between the given electrical network (1) and the non-linear and linear electrical loads (2), the output filter (11) being sized to block the harmonic components due to the switching of the inverter ( 9),

the output filter (11) is an inductance L type first order filter with a value less than 1000μH;

• a control-command unit (12) comprising a unit for calculating reference currents (25), the reference currents comprising:

• at least one non-active disturbing current intended to be injected at the connection point (C) in phase opposition, on the given electrical network side (1), the signal disturbances generated by the non-linear and linear loads (2),

exhibiting all or part of the harmonics, and at the fundamental frequency all or part of the reactive current and/or the current unbalance,

• at least one active current for recharging the capacitive energy storage element (3);

the control-command unit (12) also comprising a switching piloting device (21) which controls the switching of the inverter (9) and which provides closed-loop control of the entire frequency band from 50 to 2500 Hz for the injection by the inverter, of the non-active disturbing current and of the active current supplied via the given electrical network (1), according to the identification of the reference currents by the calculation unit (25),
the switching control of the inverter (9) orders the output filtering unit (10) to let pass at the connection point (C), part or all of the non-active disturbing currents injected in phase opposition comprising harmonic currents, as well as reactive and unbalanced currents at the fundamental frequency, in nonlinear and linear electrical loads (2),
to satisfy the non-active energy consumption demand of the non-linear and linear electrical loads (2), while depolluting the given electrical network (1) from these non-active disturbance currents. A compensation device (7) according to claim 1, wherein the switching frequency of the inverter (9) is greater than 50 kHz, preferably greater than 70 kHz, more preferably greater than 100 kHz. Compensation device (7) according to one of Claims 1 to 2, in which the first-order filter of the inductance type L has a value less than or equal to 400 μH and the switching frequency of the inverter (9) is greater or equal to 50kHz. Compensation device (7) according to one of Claims 1 to 3, in which the first order filter of the inductance type L has a value less than or equal to 200 μH and the switching frequency of the inverter (9) is greater or equal to 100 kHz. Compensation device (7) according to one of Claims 1 to 4, in which the device has only two voltage levels. Compensation device (7) according to one of Claims 1 to 5, in which the voltage upstream of the inverter (9) is greater than 800 V, advantageously greater than 1000 V. Compensation device (7) according to one of Claims 1 to 6, in which the device has a so-called digital implementation target processor for carrying out the calculations of the control-command unit (12), including that of the injected current, having a frequency lower than 100 MHz, a flash memory lower than 2 Mb and a RAM lower than 2 Mb. Compensation device (7) according to one of Claims 1 to 7, in which the device (7) operates at a fundamental frequency of between 40 and 70 Hz and the device (7) operates at a nominal voltage of the given electrical network ( 1) between 180 and 480 V. Compensation device (7) according to one of Claims 1 to 7, in which the device (7) operates at a fundamental frequency of between 360 and 800 Hz, advantageously at 400 Hz, and the device (7) operates at a nominal voltage of the given electrical network (1) between 115 and 200 V. Compensation device (7) according to one of Claims 1 to 9, in which the capacitive energy storage element (3) is connected upstream to a power unit with renewable energy generation (100),
the conversion unit (8) generating an alternating current, with a frequency band ranging from 50 to 2500 Hz, additionally covering the maximum active current generated by the renewable energy generation power unit (100),
the filter unit (10) passes the active current corresponding to the maximum power point available within the renewable energy generation power unit (100), to satisfy the active energy consumption demand of the loads electrical (2) non-linear and linear, while ensuring the recharging of the capacitive energy storage element (3). Compensation device (7) according to one of Claims 1 to 10, in which the switching control device (21) controls the switching of the inverter (9), by a controller (23) which is:

• first-order or higher-order continuous sliding mode nonlinear, or
• linear and takes into account the phase difference between the reference current and the injected current.

Compensation device (7) according to one of Claims 1 to 11, in which the unit for calculating the reference currents (25) comprises an advanced phase-locked loop unit (26), which additionally extracts from the angle provided by a conventional phase lock unit, the magnitude of the direct component of the voltage at the connection point (C) of the given electrical network (1). Compensation device (7) according to one of Claims 10 to 12, in which the switching control of the inverter (9) is implemented such that a part of the active current, in the given electrical network (1) devoid of the non-active disturbance currents, passes through the output filtering unit (10), when the production of the renewable energy generation power unit (100) is greater than the power consumed by the loads (2) nonlinear and linear electrical. Compensation device (7) according to one of Claims 10 to 13, in which the power unit with renewable energy generation (100) is coupled without a chopper or other power electronic devices to the capacitive energy storage (3). Compensation device (7) according to one of Claims 10 to 14, in which the control-command unit (12) controls:
- a chopper, configured to maintain a constant predefined DC voltage across the terminals of the capacitive energy storage element (3) of the inverter (9), independently of the voltage level of the power generation unit renewable energy (100) to ensure unchanged harmonic filtering,
- a chopper, to manage the charge/discharge of a battery bank or other energy storage systems,
- a double chopper, for the case of an island electrical network. Compensation device (7) according to one of Claims 1 to 15, in which the control-command unit (12) comprises one of the following two controllers:
- a 1 ^st order continuous sliding mode controller with an approximated sign function in Sigmoid function; Or
- an improved linear controller of the RST _am type. Compensation device (7) according to one of Claims 1 to 16, in which the switching control device (21) comprises a pulse-width modulation device (PWM), in which the command is modulated, in order to operating the inverter (9) at a fixed switching frequency and operates with the rapid switching of the power electronic components constituting the inverter. Compensation device (7) according to one of Claims 10 to 17, in which the control-command unit (12) comprises a PI-type regulator (62), with an output which represents the maximum power of the unit renewable energy generating power unit (100) for regulating the DC voltage of the capacitive energy storage element (3) while providing tracking of the maximum active power point of the renewable energy generating power unit renewable energy (100),
the compensation device (7) comprising a loop for regulating the DC voltage at the terminals of the capacitive energy storage element (3) which supplies the output of the regulator (62) with the maximum power, the voltage at the terminals of the capacitive energy storage element (3) being equal to the voltage of the maximum power of the renewable energy generation power unit (100). Compensation device (7) according to claim 18, comprising a unit for calculating the reference currents (25-A)(Upstream) configured to determine the non-active disturbing current flowing in the load (2), and a unit for calculating the reference currents (25-B) configured to calculate, based only on power, the maximum active power point voltage of the renewable energy generation power unit (100), a reference calculation unit (25-A)(Downstream) delivers the non-active/disturbing currents of the unit (25-A)(Upstream) and/or the maximum active current from the unit (25-B) and the regulator (62) , which ensures at its output the maximum power of the renewable energy generation power unit (100). Current compensation device (7) of the parallel active filter type, capable of being connected:

• at its input, downstream of at least one renewable energy generation power unit (100) coupled to a capacitive energy storage element (3), and
• in parallel, at its output, upstream to a connection point (C) between, on the one hand, a given electrical network (1), and, on the other hand, non-linear and linear electrical loads (2),

characterized in that the current compensation device (7) has:

• a power conversion unit (8), comprising at least one fast-switching voltage structure power inverter (9) with silicon carbide (SiC) or gallium nitride GaN components,

the conversion unit (8) generating an alternating current at the frequency of 50Hz, covering the maximum active current generated by the renewable energy generation power unit (100);

• an output filtering unit (10), comprising an output filter (11) for each of the phases and a neutral, and connected: on the one hand downstream of the inverter (9), and on the other hand in parallel at the connection point (C) between the given electrical network (1) and the non-linear and linear electrical loads (2), the output filter (11) being sized to block the harmonic components due to the switching of the inverter ( 9);

the output filter is an L-type inductance first-order filter and an inductance of less than 1000μH;

• a control-command unit (12) comprising a unit for calculating reference currents (25), the reference currents comprising:

• at least one active current corresponding to a maximum power point of the renewable energy generation power unit (100);
• at least one active current for recharging the capacitive energy storage element (3);

the control-command unit (12) also comprising:
- a 1st order continuous sliding mode controller (23) with an approximated sign function in Sigmoid function or an improved linear controller of RST type_am;
- a switching control device (21) which controls the switching of the inverter (9) and which provides closed-loop control at the fundamental frequency 50 Hz for the injection by the inverter of the non-active disturbing current and the active current, depending on the identification of the reference currents by the calculation unit (25);
- a PI type regulator (62), which ensures at its output the maximum power of the renewable energy generation power unit (100);
the inverter switching control (9) orders the output filtering unit (10) to pass to the connection point (C):

• an active current corresponding to a maximum power point available within the renewable energy generation power unit (100), to satisfy the active energy consumption demand of the non-linear and linear electrical loads (2), while ensuring the recharging of the capacitive energy storage element (3);
• advantageously, all or part of the currents not active at the fundamental frequency;

to satisfy the demand for reactive energy consumption of non-linear and linear electrical loads (2), while decontaminating the given electrical network (1) of these non-active disturbing currents. Compensation device (7) according to Claim 20, in which the switching frequency of the inverter is greater than 50 kHz, advantageously greater than 70 kHz, advantageously greater than 100 kHz Compensation device (7) according to one of Claims 20 to 21, in which the first-order filter of the inductance type L has a value of less than 400 μH and the switching frequency is greater than or equal to 50 kHz. Compensation device (7) according to one of Claims 20 to 22, in which the voltage upstream of the inverter (9) is greater than 800 V, advantageously greater than 1000 V. Compensation device (7) according to one of Claims 20 to 23, in which the power unit with renewable energy generation (100) is also coupled without a chopper or other power electronic devices to the element capacitive energy storage (3). Compensation device (7) according to one of Claims 20 to 24, in which the control-command unit (12) controls:
- a chopper, configured to maintain a constant predefined DC voltage across the terminals of the capacitive energy storage element (3) of the inverter (9), independently of the voltage level of the power generation unit renewable energy (100) to ensure unchanged harmonic filtering,
- a chopper, to manage the charge/discharge of a battery bank or other energy storage systems,
- a double chopper, for the case of an island electrical network. Electrical system comprising a given electrical network (1), non-linear/linear loads (2), a compensation device (7) according to one of the preceding claims 1 to 25. A system according to claim 26, wherein the system is connected upstream to a renewable energy generating power unit (100) which is:

• chosen from the following list: one or more photovoltaic panels, wind turbine(s), fuel cell(s), and
• coupled, directly or via a DC/DC power electronics device to a capacitive energy storage element (3) in the case of continuous production, or via an AC/DC power rectifier in the case of production alternative.

System according to one of Claims 26 to 27, in which the given electrical network (1) is chosen from the following list: the main electrical network, a local electrical micro-grid islanded or connected to the main electrical network, or an electrical network embarked. System according to one of Claims 26 to 28, further comprising an intelligent building (27), and in which the control-command unit (12) is connected to a decentralized management unit (70) of the intelligent building (27 ). System according to claim 29, in which the control-command unit (12) is configured to optimize the consumption of the various devices operating within this intelligent building (27) by distributing the loads corresponding to the non-linear/linear loads according to at least two operating modes:

• a first distribution mode, called adapted consumption mode, in which the control-command unit (12) controls the decentralized management unit (70) of the intelligent building (27) so as to adapt the consumption of the intelligent building with producing the renewable energy generating power unit (100) so that the total load curve of the intelligent building has a maximum simultaneity factor corresponding to the operation of all the payloads of the building at the same time, in the limit of renewable energy produced,

the first mode having charge/discharge management of the batteries or other energy storage systems within the same positive energy smart building (27) or between the buildings (27) interconnected via the control-command units (12 ) devices (7);

• in the event of insufficient renewable energy production, a second distribution mode, called modulated consumption mode, in which the control-command unit (12) controls the decentralized management unit (70) of the intelligent building ( 27) so as to modulate the consumption of the devices of the intelligent building (27) to tend to a substantially constant total load curve of the intelligent building as a function of time.

System according to one of claims 26 to 30, further comprising:

• a local network (28) connected to the given electrical network (1), and
• conventional power generation units (80), and
• a semi-decentralized management system (29), and
• a plurality of renewable energy generation power units (100) connected to the local network (28), by a compensation device (7), each compensation device (7) being connected to the semi-decentralized management system ( 29) to which it communicates information concerning the current and future energy production of each of the renewable energy generating power units (100), and
• a plurality of consumption stations corresponding to the non-linear and linear loads (2), each of the consumption stations being connected to the local network (28) and equipped with a compensation device (7), connected to the semi-decentralized management system (29) to which it communicates information concerning the instantaneous consumption and the consumption to come according to the programmed operation of the consumption stations,
• a plurality of renewable energy generation power units (100) and linear and non-linear loads (2) of positive energy smart buildings (27), which ensure self-consumption and where any excess energy is stored or exchanged with the other intelligent buildings or delivered to the local network (28) via the control-command unit (12) of the compensation device (7), in coordination with the semi-decentralized management system (29) in the event exchange with the local network (28).

System according to one of Claims 26 to 31, in which the semi-decentralized management system (29) is configured to control the distribution on the local network (28) of the power coming from the given electrical network (1) if the production estimated total does not cover demand; or to intervene when the total energy demand within the local network (28) is greater than the total production, with the decentralized management units (70) of the intelligent buildings (27), via the control-command units (12 ) compensation devices (7), to make them switch to a modulated consumption mode.