FR3044174A1 - SYSTEM AND METHOD FOR REGULATION OF VOLTAGE AND FREQUENCY OF AN ISOLABLE NETWORK - Google Patents

Abstract

The invention relates to a system for regulating (3) an amplitude and a frequency of a voltage delivered on a network (1) that can be isolated from a main network (2), and comprising a device for generating non-regulated electricity (10) and a consumption load (12), characterized in that it comprises: - two regulating charges (30, 31) adapted to consume or respectively produce a variable active and reactive power respectively as a function of an active power setpoint value and a reactive power setpoint value, a control unit (33), comprising: ○ at least one sensor (34) for measuring the amplitude and frequency of the delivered voltage to the consumption load, ○ a processor (35) adapted, from the measurements of the sensors, to deliver to the control loads the set values of active power and reactive power.

Description

FIELD OF THE INVENTION The invention relates to a system for regulating the amplitude and the frequency of the electric voltage delivered on a network comprising an unregulated power-generating device, and which can be isolated from an upstream network provided with regulated power generation device. The invention also relates to a control method implemented with such a system.

STATE OF THE ART

Some isolated locations, such as mountainous areas or islands, have power grids that may be isolated from a public grid (or backbone) in certain circumstances, such as in difficult weather conditions incident on the main power grid.

In isolation, local energy resources can be used to maintain power generation in the isolated network. However, such a network must have a voltage whose amplitude and frequency are regulated.

Some sources of electricity production are able to produce a voltage whose amplitude and frequency are regulated; this is the case, for example, of generators equipped with synchronous generators and voltage and frequency regulators, or sets comprising photovoltaic panels, batteries, inverters, and voltage and frequency regulators.

These means of production are expensive, so most electricity distribution networks do not have them.

On the other hand, many distribution networks are equipped with unregulated power generation means, for example wind turbines, photovoltaic panels without control systems, and so on. However, these means do not make it possible to meet the requirements of the electrical consumption loads concerning the amplitude and the frequency of the electrical voltage delivered on the network.

In addition, these means of production can produce energy only after having been connected to an already undervoltage network, that is to say that they are not themselves capable of powering a power supply. network initially off.

To solve this problem, for example, document FR 2 437 105 describes a means of initiating the operation of a generator by capacitor banks. However, this document does not make it possible to regulate the amplitude and the frequency of the voltage after the priming.

Also known are documents FR 2 481 857, FR 2 870 401 and FR 2 490 421 which describe means for regulating the amplitude and the frequency of the voltage of an isolated network, the amplitude of the voltage being regulated by the control of reactive charges, and the frequency being controlled by controlling the speed of a drive motor of the generating generator producing electricity.

These documents do not solve the problem of the priming of the isolated network, and moreover they require to double the generator of electricity by a motor whose speed is controlled to regulate the frequency.

PRESENTATION OF THE INVENTION The purpose of the invention is to overcome the disadvantages of the prior art described above.

In particular, an object of the invention is to propose a system for regulating the amplitude and the frequency of a voltage delivered on an autonomous network, which allows both the initiation of the network as well as its regulation in regime. stabilized.

Another object of the invention is to enable the frequency of the voltage to be regulated without the need for a motor to drive the electricity generating device. In this regard, the subject of the invention is a system for regulating an amplitude and a frequency of an electric voltage delivered on a network comprising an unregulated electricity generating device and at least one consumption charge. electricity, said network being capable of being isolated from a main network, the system being characterized in that it comprises: a first regulation load adapted to consume or produce a variable active power as a function of a value; active power setpoint, a second regulation load, adapted to consume or produce a variable reactive power as a function of a reactive power setpoint, and a control unit, comprising: at least one sensor adapted to measure: the amplitude of the voltage delivered to the consumption load, the frequency of the voltage delivered to the consumption load, o a processor ad suitable for measuring the active and reactive power setpoints from the sensor measurements.

Advantageously, but optionally, the system according to the invention may further comprise at least one of the following features: the first regulating load comprises at least one variable resistor, and the second regulating load comprises at least one capacitor of variable capacitance .

The control system further comprises a load adapted to consume or produce a fixed reactive power.

The processor is further adapted to implement a plurality of control laws of the active power and reactive power setpoint values, and to select a control law to be implemented according to an initial state of the network among the following group: o the network is de-energized and isolated from the upstream network, o the network is live and isolated from the upstream network, and o the network is powered up and connected to the upstream network, the control laws include at least one control determining the active power and reactive power setpoint values from the measurements of the amplitude of the voltage delivered to the consumption load and the frequency of the voltage delivered to the consumption load.

The system further comprises a load adapted to consume or produce a fixed reactive power, wherein the control laws comprise at least one control law determining the active power and reactive power setpoint values from a phase shift measurement. between the voltage and the current and from a measurement of voltage amplitude or voltage frequency delivered to the consumption load. The subject of the invention is also a method of regulating an amplitude and a frequency of an electric voltage delivered on a network comprising an unregulated electricity generating device and at least one electricity consumption load, said network being capable of being isolated from a main network, the method being characterized in that it is implemented by a control system according to the foregoing description, and in that it comprises: - the detection of an initial state of the network, from the initial state of the detected network, the selection of at least one control law of reactive power setpoint values and of active power to be applied to the regulation loads to regulate the amplitude and frequency of the electrical voltage on the network.

Advantageously, but optionally, the control method according to the invention may further comprise at least one of the following features: the method is implemented in a control system further comprising a generator, and, if the state network is de-energized and isolated from the main grid, the process includes: o powering up the grid by the generator, o control of active and reactive power loads at fixed setpoints, o connection of the device unregulated power generation, and o the disconnection of the generating set and the control of the active and reactive power loads according to a steady-state regulation control law.

The method is implemented in a control system comprising an islanding contactor adapted to selectively connect or disconnect the network to the main network, and wherein, if the initial state of the network is undervoltage and connected to the upstream network, the method comprises: the control of the active and reactive power loads at values of setpoints respectively function of the active power and the reactive power passing through the islanding contactor, the opening of the islanding contactor, and the control of active and reactive power loads according to a regulation law in steady state. - If the initial state of the network is undervoltage and disconnected from the upstream network, the method comprises controlling the active and reactive power loads according to a steady state control law. The invention also relates to a control method implemented by such a system.

The proposed control system includes a regulating load that can consume or produce variable active power, a regulating load that can consume or produce variable reactive power, and a control unit that can control these regulating loads to regulate both the amplitude of the voltage and the frequency of the voltage delivered on the autonomous network.

This system can regulate the amplitude and frequency of the voltage in different states in which the network may be. In particular, it makes it possible to initiate the powering on of the network when it is initially disconnected from a main network and off, or it also makes it possible to isolate the isolable network of the main network, without generating a power cut. .

This system also makes it possible to provide steady-state regulation, in which the network is isolated from the main network and yet energized, to meet the requirements for consumption of loads such as domestic electrical equipment.

In addition, the control unit can, when the network is isolated from the main network and under tension, implement different control laws of the set values to be given to the regulation loads, according to pre-recorded scenarios. This makes it possible to choose the optimal control law, as a function, for example, of the technologies selected for the regulation loads and the requirements of the consumption load on the voltage and the frequency in terms of accuracy and speed of regulation.

DESCRIPTION OF THE FIGURES Other characteristics, objects and advantages of the present invention will appear on reading the detailed description which follows, with reference to the appended figures, given by way of non-limiting examples and in which:

FIGS. 1a and 1b show a regulation system according to two embodiments of the invention,

FIGS. 2a and 2b show equivalent electrical diagrams of a control system according to two variant embodiments of the invention,

Figure 3 shows schematically the main steps of a control method implemented by the system.

DETAILED DESCRIPTION OF AT LEAST ONE EMBODIMENT OF THE INVENTION

Regulation system

With reference to FIGS. 1a and 1b, there is shown diagrammatically a control system 3 of an isolable network 1, also called a micro-network or microgrid, with unregulated power generation means.

The isolable network 1 comprises at least one unregulated electricity generator 10, which may be of the photovoltaic panel, wind turbine, gas generator, etc. type. No generator 10 is provided with regulation means dedicated to controlling the amplitude or the frequency of the voltage at its terminals. The generators 10 are therefore normally provided to supply electricity to a main network 2 or upstream network, which in turn has regulation means and regulated electricity generation means.

The isolable network may also include auxiliary electric charges, which are typically the auxiliaries of the generators 10, such as pumps belonging to a gas generator, or the control equipment required for its operation. To start the generators 10 it is therefore necessary to give these loads 11 the active and reactive powers they require.

Finally, the network 1 may also include an electrical generator 12 adapted to power up the network when it is initially off, and maintain the frequency and amplitude of the voltage within acceptable ranges of values. Typically this generator is a generator 12.

The generators 10, 12 and the auxiliary charges 11 are preferably grouped together in a so-called central zone Z.

The isolable network 1 also comprises consumption charges 13 arranged in a so-called consumption zone Z ', typically particular customers having domestic equipment. These charges are adapted to consume active power and to consume or produce reactive power.

The central zone Z is electrically connected to the consumption zone Z 'by an interzone electrical link 14. In some cases, as shown in FIG. 1b, the consumption area Z' may also comprise an unregulated electricity generator 10. This does not prevent the regulation described below.

In addition, when the isolatable network 1 can be selectively connected to the upstream network 2 (shown in FIG. 1a) or disconnected from it, this selective connection is made by an islanding contactor 15 (shown in FIG. 1a).

The regulation system 3 makes it possible to regulate the amplitude and the frequency of the voltage delivered by the generator (s) 10 of the network 1. For this purpose, it comprises a first regulating load 30, adapted to consume or produce a variable active power .

This regulation load 30 may comprise for example one or more resistors, the resistances being variable and / or connected to the network 1 by means of a power electronics or a transformer with on-load tap changer. The regulating load 30 may also be an inverter battery system.

The active power consumed or produced by this first regulation load can be controlled with a setpoint ΔΡ.

The regulation system 3 comprises a second regulating load 31, adapted to consume or produce a variable reactive power. This load 31 may comprise, for example, one or more capacitors of variable capacitance, power electronics components such as inverters, or a synchronous compensator.

The reactive power consumed or produced by this second regulation load 31 can be controlled with an AQ setpoint.

Advantageously, but optionally, the control system may further comprise a fixed reactive load 32, consuming or providing a fixed reactive power Qf.

The control system further comprises a control unit 33, adapted to deliver to the regulating loads 30, 31 the appropriate setpoint values on the active power and the reactive power in order to regulate the amplitude and the frequency of the voltage on the network. In this respect, the control unit 33 comprises at least one sensor 34, adapted to measure: the amplitude U of the voltage delivered by the generator 10 to the consumption load 13, and the frequency f of the voltage delivered by the generator 10 to the consumer load 13.

As can be seen in Figures 1a and 1b, these measurements can be performed at the interzone electrical link 14.

Advantageously, the sensor or sensors 34 of the control unit 33 are also adapted to measure the active power P, and the reactive power Qi passing through the islanding contactor 15.

Advantageously, in the case where the network 1 comprises a generator 12, the sensors 34 of the control unit are also adapted to measure the active power PGe and the reactive power QGE produced by the group 12.

In addition, the control unit 33 comprises processing means, typically a processor 35, and a memory 36, shown only in Figure 1a. The processor 35 is adapted to receive the measurements of the sensors 34 and, from these measurements, to send to the control loads 30, 31, setpoints ΔΡ, AQ on the active power and the reactive power in order to regulate the amplitude and frequency of the voltage on the network.

To do this, the processor 35 is adapted to selectively execute a plurality of control laws A, B, C, D, E or F, which are stored in the memory 36. Each control law determines a set value respectively for the active power ΔΡ, and for the reactive power ΔΟ according to respective input parameters a and b separately.

We have AP = f (a) and AQ = f (b).

The control laws A to F have different input parameters which will be described in more detail below.

The control law executed among the various available laws is chosen by the processor according to the initial state of the network 1 as well as the circumstances of the regulation.

The network 1 may be in one of the following initial states: it is off-line and disconnected from the main network; the islanding contactor 15 is open. This is typically the case when the network connection 1 is accidentally cut off from the main network 2.

The network 1 is under-voltage and connected to the main network 2; the island connector 15 is closed. This is the case when the network 1 is still powered by the main network and it is desired to disconnect it from this network without generating a power cut. - Network 1 is under-voltage and disconnected from the main network; the islanding contactor 15 is open. This is the case when the generation of electricity has been started by the generator 12 to restore the voltage after a cutoff of the network connection 1 to the main network 2.

With regard to the first two initial states described above, the control system makes it possible to pass the network 1 of one of these initial states to a stabilized state in which the network 1 is under-voltage t disconnected from the main network, which corresponds to the third state described above. Then the control system 3 is adapted to maintain the network in this state by regulating the frequency and amplitude of the voltage to meet the requirements of the electrical charges. In this respect, the control laws comprise four steady-state control laws A, B, C, D, allowing the system to regulate the amplitude and frequency of the voltage on the network once it is energized. , and two transient state control laws E and F, allowing the system to bring the network from an initial state to a state in which it is energized and disconnected from the main network. The details of the control laws are described in more detail below.

Finally, the processor 35 is advantageously adapted to control actuators (not shown) to open and close the islanding contactor and open and close contactors for connecting the generators 10, the generator 12, and the regulating loads 30, 31 to the network 1.

Regulation process

With reference to FIG. 3, the main steps of the regulation method implemented by the system 3 above are schematically represented.

The method comprises a first step 100 of detection of an initial state of the network by the control unit 33. This step is implemented for example by measuring the amplitude of the voltage on the network 1 and by measuring the values of the active and reactive power P | and Q, passing through the islanding contactor 15.

If the voltage in the network is zero, then the initial state is the off state 110, disconnected from the main network 2.

If the voltage in the network is non-zero, and a non-zero active and / or reactive power passes through the islanding contactor, then the initial state is the state 120 under voltage and connected to the main network 2.

If the voltage in the network is non-zero, but the active and reactive power passing through the islanding contactor is zero, then the initial state is the state 130 under voltage and disconnected from the main network.

The method then comprises a regulation step 200 as a function of the initial state of the network 1 detected in the previous step.

When the initial state is the off state 110 and disconnected from the main network 2, the network 1 must be re-energized before being stabilized. Consequently, the regulation step 200 comprises a first step 210 for connecting the generator set 12 to the grid and for starting the generator set so that it can power up the network 1.

This brings the amplitude and the frequency of the voltage delivered by the generator set to determined values. As a result, the generator set 12 generates active power PGe and the reactive power Qge-

To then connect the unregulated power generation sources 10 to the network, the control unit 33 connects the active and reactive regulation loads 31 to the network, and the processor implements a control law E in which the active and reactive power reference values ΔΡ, ΔΟ are fixed and independent of the other quantities of the network. In other words, this control law has no input parameters a and b. The control unit 33 then connects the unregulated electricity generator (s) to the grid and turns them on. In this way, the generators produce a growing active power.

To maintain the balance of the active powers produced and consumed on the network 1, the generator will decrease the active power it produces. Once the active power produced by the generator passes below a predetermined threshold stored in the memory 36, the control unit disconnects 214 the generator.

The network is in the undervoltage state and disconnected from the main network 2, which corresponds to the initial state 130 introduced above. The amplitude and the frequency of the voltage are then regulated by the implementation of one of the control laws A, B, C, or D described in more detail below.

When the initial state of the network 1 is the state 120 under voltage and connected to the main network 2, the network 1 must be disconnected from the main network 2 without generating a cutoff. The regulation step 200 comprises a first step 220 for connecting the active and reactive regulation loads 31 to the network by the control of their respective contactors.

The processor 35 of the control unit then implements 221 a control law F. This control law delivers the active and reactive power setpoints ΔΡ, ΔΟ as a function of the active power values P | and reactive Qi passing through the islanding contactor 15 and measured by the sensors 34. The control law adjusts the values of the setpoints to converge the active power P | and reactivates Qi crossing the island contactor to 0.

For this law F we have ΔΡ = ί (Ρ |) and AQ = f (Qi)

For example, the control law can increase the active power consumed by the regulating load 30 to reduce or increase the active power passing through the islanding contactor 15, depending on the direction in which the active power passes through the islanding contactor.

When the control unit 33 detects, by means of the sensors 34, the moment when the active power Pi and reactive Q, passing through the islanding contactor passes under a predetermined value, it controls 222 the islanding contactor 15 to isolate the island. network 1 of the main network 2.

The network is then in the energized state and isolated from the main network 2 which corresponds to the initial state 130 introduced above.

Finally, when the state of the network 1 corresponds to the state under voltage and isolated from the main network 2 - whether this is an initial state or a state reached after the implementation of the control laws E or F, the processor 35 of the control unit 33 is adapted to implement 230 a control law among the laws A, B, C, and D. Each of these laws aims to regulate the amplitude and frequency of the voltage delivered by the Unregulated electricity generator 10, i.e. converging the amplitude of the voltage U to a predetermined value Uo and converging the frequency f to a predetermined value f0.

To achieve these convergences, each control law modifies the setpoints ΔΡ and ΔΟ, the modification of each setpoint causing variations of the parameters U, f and a, with the phase shift between the voltage and the current. At a given moment, a single control law is implemented by the processor 35.

Table 1 shows a summary of the control laws A to F, and the parameters from which these laws generate setpoints on the active power and the reactive power.

Table 1

Thus, for the law A, the processor 35 calculates the setpoint ΔΡ as a function of U and the setpoint ΔΟ as a function of f.

For the law B, the processor 35 calculates the setpoint ΔΡ as a function of f and the setpoint AQ as a function of U.

For the law C, the processor 35 calculates the setpoint ΔΡ as a function of U and the setpoint Δ0 as a function of a.

For the law D, the processor 35 calculates the setpoint ΔΡ as a function of f and the setpoint Δ0 as a function of a.

Advantageously, the processor 35 selects a law from among the laws A to D of stabilized speed according to pre-recorded scenarios in the memory 36, taking into account parameters such as: - Structure and operation of the active and reactive control charges 31, and

Requirements of consumption loads 13 on the accuracy of regulation, in amplitude of voltage or in frequency.

For example, if the reactive regulation load 31 is formed by a step of capacitors, it does not make it possible to fine-adjust the capacitance and therefore the reactive power that it supplies or consumes, and also has constraints on time. re-engagement of the capacitors.

If the active regulation load 30 is based on power electronics, it allows fine adjustment of the resistance and therefore the active power that it consumes, and very quickly.

Therefore, if the consumption load 13 is very demanding on the amplitude of the voltage and less on the frequency, the law A is more appropriate than the law B because it makes it possible to respond quickly to voltage variations and thus to demonstrate a greater precision.

Referring to Figures 2a and 2b, we will now detail the principle of implementation of laws A and B on the one hand, and laws C and D on the other hand, in the case where the generator 10 is an asynchronous generator .

In FIG. 2a, there is shown an equivalent electrical diagram of the isolated network 2, where R1, L1 and Lee represent the generator (s) 10: R1 is a negative resistance, it represents the production of active power, L1 is a coil, it represents the reactive power consumption of the generators - in the case of asynchronous generators it is the magnetising inductance,

Lee is the short-circuit impedance of the generators. R2 and L2 represent the consumption loads 13: R2 is a variable resistance depending on the electricity consumption of a customer, L2 is a coil whose inductance value varies according to the electricity consumption of a customer .

Finally, R is a resistance that represents the active power control load. The set point ΔΡ consists of changing the resistance value of R. C is a capacitor representing the reactive power control charge 31. The set point ΔΟ consists of changing the value of the capacitance of C. U is the amplitude of the voltage delivered to the consumption load 13.

In such a circuit, the equilibrium voltage Ueq and the resonance frequency feq depend on the values of the resistance of R and the capacitance of C.

Thus, the control laws A and B drive the set values to adjust the resistance values of R and capacitance of C to maintain Ueq at the desired Uo, typically 400V, and feq at the desired f0, typically 50Hz.

The control laws vary depending on the network and the nature of the regulatory loads.

In order to implement the control laws A and B, the positions of the control loads 30 and 31 are arbitrary. The control system may optionally include a fixed reactive load. In particular, the embodiments shown in FIGS. 1a and 1b make it possible to implement these control laws.

Thus according to a non-limiting example, as in Figure 1a, all these charges can be connected to the network 1 in the central area. Alternatively, shown in Figure 1b: at least one load can be connected to the network in the central zone Z, for example the fixed load 32, at least one load can be connected to the network in the consumption zone Z ', by For example, the reactive control load 31, and at least one load can be connected to the network at the interzone electrical link 12, for example the first control load 30.

With reference to FIG. 2b, the equivalent electrical diagram of the insulated network 2 is again shown, in the case where the network also comprises a fixed reactive load 32 represented by a capacitor C1.

The subset consisting of the capacitor C1, the coil L1 and Lee coil, and defined by dashed lines in Figure 2b, is a resonant subset. The capacitance of the capacitor, i.e. the reactive power supplied or absorbed by the fixed load 32, is set so that the subset resonance frequency is equal to a predetermined frequency f0. a being the phase shift between the voltage U and the current I in the network, provided, when it is not zero, information on the level of production or reactive power consumption of the consumption zone Z '. When at zero, the reactive powers produced by the regulating load 31 (capacitor C) and consumed by the consumption load are identical.

Thus, the control laws C and D drive the setpoint value reactive power AQ as a function of the measured value of the phase shift a, to maintain a to 0.

In the control law C, the active power setpoint ΔΡ is a change in the value of the resistance of R as a function of the measured value of the frequency f to reduce this frequency to a predetermined value f0.

In the control law D, the active power setpoint ΔΡ is a change in the value of the resistance of R as a function of the measured value of the amplitude of the voltage to reduce this amplitude to a predetermined level U0.

In order to be able to implement the control laws C and D, the position of the regulation loads 30, 31 must respect the following constraints: the reactive power regulation load 31 must be connected to the network in the consumption zone Z ', the reactive fixed load 32 must be connected to the network in the central zone Z,

Finally, there must be an active power transit between the central zone Z and the consumption zone Z '.

These constraints are respected in the illustration of Figure 1b.

The proposed control system therefore makes it possible to operate a transient isolable network to a steady state in which the network is switched on and disconnected from a main network. In addition, the regulation of the voltage and the frequency under steady state can be implemented according to different control laws, the most relevant control law being able to be chosen as a function of the structure of the network, the type of regulation loads and load requirements.

Claims (10)

Control system (3) for an amplitude and a frequency of an electrical voltage delivered on a network (1) comprising an unregulated electricity generating device (10) and at least one consumption charge d electricity (12), said network being capable of being isolated from a main network (2), the system (3) being characterized in that it comprises: a first regulating charge (30) adapted to consume or producing a variable active power as a function of an active power setpoint value, a second regulating load (31) adapted to consume or produce a variable reactive power as a function of a reactive power setpoint, and a unit control unit (33), comprising: o at least one sensor (34) adapted to measure: the amplitude of the voltage delivered to the consumption load, the frequency of the voltage delivered to the consumption load, o a processor (35) adapted, from the measurements of the sensors, to deliver to the control loads (30, 31) setpoint values of active power and reactive power. 2. Control system (3) according to claim 1, wherein the first regulating load (30) comprises at least one variable resistor, and the second regulating load (31) comprises at least one capacitor of variable capacitance. 3. Control system (3) according to one of claims 1 or 2, further comprising a load (32) adapted to consume or produce a fixed reactive power. 4. Control system (3) according to one of the preceding claims, wherein the processor (35) is further adapted to implement a plurality of control laws of the set values of active power and reactive power, and to select a control law to be implemented according to an initial state of the network among the following group: the network is de-energized and isolated from the upstream network, the network is under voltage and isolated from the upstream network, and the network is energized and connected to the upstream network. 5. Control system (3) according to claim 4, and wherein the control laws comprise at least one control law determining the active power and reactive power setpoint values from the measurements the amplitude of the voltage delivered. the consumption load (13) and the frequency of the voltage delivered to the consumption load (13). 6. Control system (3) according to one of claims 4 or 5, further comprising a load (32) adapted to consume or produce a fixed reactive power, wherein the control laws comprise at least one control law determining the active power and reactive power setpoint values from a phase shift measurement between the voltage and the current and from a measurement of the voltage amplitude or voltage frequency delivered to the consumption load (13). ). A method of controlling an amplitude and frequency of an electrical voltage delivered on a network (1) comprising an unregulated power generating device (10) and at least one power consumption load ( 13), said network being capable of being isolated from a main network (2), the method being characterized in that it is implemented by a control system (3) according to one of the preceding claims, and in that it comprises: detecting (100) an initial state of the network, from the initial state of the detected network, selecting at least one control law for reactive power setpoint values and active power to be applied to the control loads to regulate (200) the magnitude and frequency of the mains voltage on the network. 8. Control method according to claim 7, implemented in a control system (3) further comprising a generator (12), and wherein, if the initial state of the network is off and isolated from the main network (2), the method comprises: - energizing the network by the generator set (210), controlling the active and reactive power loads to fixed set point values (211,212), - connecting the production device to unregulated electricity (213), and - the disconnection of the generator set (214) and the control (230) of the active and reactive power loads according to a stabilized regulation control law. 9. Control method according to one of claims 7 or 8, implemented in a control system (3) comprising an islanding contactor (15) adapted to selectively connect or disconnect the network (1) to the main network ( 2), and wherein, if the initial state of the network is undervoltage and connected to the upstream network, the method comprises: controlling the active and reactive power loads at setpoint values respectively related to the active power and the reactive power passing through the islanding contactor (221), the opening of the islanding contactor (222), and the control of the active and reactive power loads according to a steady-state regulation law (230). 10. Control method according to one of claims 7 to 9, wherein, if the initial state of the network is undervoltage and disconnected from the upstream network, the method comprises controlling the active and reactive power loads according to a law stabilized control system (230).