RO130090B1 - Method for decoupling the oscillating power for single-phase inverters - Google Patents

Description

The invention relates to a method of decoupling the oscillating power for single-phase inverters, in order to attenuate the alternating current component at the input of the inverter and to reduce the capacitor capacity required on the direct current line of the inverter. The invention falls within the field of electronic power converters for dc-dc conversion. and c.a.-c.c., as well as the control associated with them.

Several solutions are known for attenuating the alternating power component on the DC line of single-phase inverters. Currently, the passive filtration solution, which uses a high capacity electrolytic capacitor on the d.c. line. of the inverter, predominates in all applications due to its simplicity. However, new solutions are being sought to eliminate the main disadvantage of the passive method, namely the short life of the electrolytic capacitor which affects the reliability of the entire system as described in the paper “Reliability of Candidate PhotovoltaicModule-lntegrated-lnverter (PV- MII) Topologies-A Usage Model Approach ”, S. Hârb, RS Balog, IEEE Transactions on Power Electronics, vol. 28, no.6, 2013, pp.3019-3027.

Existing active decoupling methods are based on additional circuits operating as active power filters that extract the oscillating power component and divert it to a short-term storage element that may be a small-capacity capacitor, or a coil. Low capacity polypropylene film capacitors are preferred due to their long service life, thus eliminating the aforementioned problem with electrolytic capacitors. However, existing active solutions require additional components, such as: transistors, diodes, filters. Therefore, the complexity of converters will increase, their cost will be higher and the overall efficiency will decrease. A recent analysis of the main existing decoupling methods, mainly those for micro-inverters for photovoltaic (PV) panels, is presented in "A Review of Power Decoupling Techniques for Micro-inverters with Three Different Decoupling Capacitor Locations in PV systems". H. Hu, S. Hârb, N. Kutkut, I. Batarseh, ZJ Shen, IEEE Transactions on Power Electronics, vol. 28, no. 6, 2013 pp. 2711 - 2726. It is found that the solutions analyzed in the technical literature require at least one additional semiconductor element and passive elements. Next, existing inventions worldwide will be discussed, highlighting their main disadvantages compared to the proposed invention.

A general solution for minimizing the alternative power component in single-phase converters is presented in U.S. Patent 83,25499 B2, “Methods for minimizing doublefrequency ripple power in single-phase power conditioners.” Philip T. Krein, Robert S. Balog, 2012. This consists of in the addition of an energy storage system and a power interface to divert the oscillating power component, which is transferred from a source to a load, or single-phase network, to the energy storage unit (eg capacitor). In that patent, a control method is proposed, which is based on the generation of a voltage on the storage element out of phase with 45 ° to the AC output voltage. The method involves the use of an additional bidirectional converter for the storage element, which is a disadvantage of the invention.

in US patent 2011261593A1 “Power converter for a power generator”, Shangzhi Pan, Sayed Aii Khajehoddin, Praveen K. Jain, Alireza Bakhshai, 2011, an electronic power converter is proposed for interfacing low power generators (eg PV, turbines wind) with the single-phase network. The converter contains two power conversion stages, at the input a d.c.-d.c. stage, respectively an inverter at the output. By controlling the two floors, the elimination of the oscillating component from the input is obtained, which will be processed on the d.c. line. of the inverter where a capacitor of low capacity is used (two orders of magnitude smaller than in the case of the passive filtering method). As a result, the dc voltage. it will not remain constant

RO 130090 Β1 (or with negligible ripple) as in the case of conventional inverters, but will contain a high alternative component 1. To eliminate the harmonic pollution of the current injected in the network as a result of the variation of the dc voltage. a correction is applied to the reference voltage applied to the PWM 3 generator for the inverter control. Although the invention ensures the decoupling of the alternative power component by using a capacitor of low capacity, its main disadvantage 5 is that it can be applied only by adding a c.c.-c.c. conversion stage. Also, for applications that require bidirectional power transfer between input and output 7 (eg in the case of an energy storage system), the complexity of the d.c.-d.c converter. will increase considerably. 9

Another invention, US 2012257429A1, “Two-stage single phase bi-directional PWM power converter with DC link capacitor reduction”, Dong Dong, Dushan Boroyevich, Ruxi Wang, Fred 11 Wang, 2012, can provide substantial reduction of the filter capacitor cc, can also be adapted for applications with bidirectional power transfer, and consists in adding an arm 13 with two series transistors with active filter role on the cc line of the inverter. This ensures that the power oscillating component is diverted to this capacitor. An application in 15 is exemplified in which the reduction of the capacitor capacity is about 30 times compared to the passive filtration. The main disadvantage of the method is the addition of two transistors and a filter inductor 17 to ensure the decoupling function, increasing the complexity of the system which will be reflected in a higher cost. 19 in US patent 8670249 B2, “Inverter for a distributed power generator”, Sayed Aii Khajehoddin, Praveen Jain, Alireza Bakhshai, 2014, a current inverter is proposed for 21 connecting the PV panels to the single-phase network, in which a power inductor is used. cc if a c.c.-c.c. with a transistor and a diode that ensures, in addition to adapting the input voltage level 23, the decoupling function. In this case no capacitors are used but the coil within the current inverter is used as a short-term storage element 25 for filtering the oscillating power component. The coil will be traversed by the direct load current to which a high alternating component is added, which can lead to saturation 27 of the magnetic circuit. In addition to this aspect, the decoupling additionally requires a transistor and a diode, which are the main disadvantages of said invention. 29

The invention aims to develop a method for decoupling the oscillating power component for single-phase H-bridge inverters, without using additional semiconductor elements.

The technical problem solved by the invention consists in ensuring the decoupling of the oscillating component of the power in the case of single-phase inverters.

This technical problem is solved only by controlling the inverter and by using 35 non-polarized capacitors with polypropylene film, connected on each arm of the inverter bridge between the midpoint and one of the terminals of the dc line. 37

The method of decoupling the oscillating power for single-phase inverters, which ensures the decoupling of the oscillating component of the instantaneous power, is applied to a circuit consisting of a 39 H-bridge inverter, controlled by width-modulated pulses (PWM), which is supplied by a line of direct current, the midpoints of the two arms of the inverter bridge being 41 connected to two filter coils, and to the output terminals of the coils being connected two non-polarized capacitors of ca designed to disconnect the oscillating component of the instantaneous output power 43, the other two terminals of the capacitors being connected together to one of the dc, DC + or DC- terminals of the inverter, and the differential output of the inverter 45 can supply a load and the method consists in:

- generation of two voltages u _AB and u _comp by a control structure consisting of two 47 main subsystems, a subsystem providing the reference value of the differential output voltage u _AB and another subsystem providing a calculated compensation voltage u ^ p in office 49

EN 130090 Β1 of the apparent power S of the load at the output of the inverter, the pulsation of the output voltage ω, the phase shift angle φ between the voltage and the load current and the capacity of the disconnect capacitors C _d

- multiplying the signal u _AB by a factor of 1/2 in an amplification block, the resulting signal is collected by means of an adder with the signal u _comp , resulting in a control voltage u _A , which is divided by means of a division block at the voltage of on the DC line _DC , provides a signal with the filling factor D _A for the first arm of the inverter,

- the decrease of the signal u _AB multiplied by a factor 1/2 in the amplification block, from the voltage u _comp in the adder, resulting in a control voltage u _B , which divided by means of a division block at the voltage on the DC line U _DC , provides a signal with the filling factor D _B for the second arm of the inverter,

- control with the two signals D _A and D _B of the transistors T _r T ₄ within the H-bridge of the inverter by means of a PWM pulse generator, thus ensuring the decoupling of the oscillating component of the instantaneous power of the inverter.

The main advantage of the proposed invention is that it provides the decoupling function under similar conditions to the solutions presented above, but without the use of additional semiconductor elements.

Another advantage of the invention is that it can be used in various applications requiring a single-phase inverter, which can operate unidirectional (eg interfacing renewable energy sources to the grid, or powering a single-phase load), or bidirectional (eg energy storage). Due to the small number of components required, the complexity of the system by the applied method is reduced, which is another advantage of the invention.

An embodiment of the invention is given below in connection with figures 1 ... 3, which show:

- fig. 1, the diagram of the single-phase inverter with the proposed decoupling method;

- fig. 2, the main waveforms characteristic of the single-phase inverter with the decoupling method according to the invention;

- fig. 3, experimental results:

a) i _dc with and without decoupling;

b) U _dc with and without decoupling;

c) harmonic analysis of the i _dc current with and without decoupling;

d) the voltage (u _Cd1 ) and current (i _Cd1 ) of a decoupling capacitor;

e) voltage (u ₀ ) and current (/ ₀ ) on load, current through the first arm of the inverter (i _A ).

FIG. 1 shows the diagram of the single-phase inverter with a decoupling circuit of the alternative power component and the control structure that ensures the proposed decoupling function. The system consists of a voltage inverter 1 in the H bridge with four transistors T1-T4, which is supplied by a DC line 2 with the voltage U _dc . T1-T4 transistors are controlled by width modulated pulses (PWM). The midpoints A and B of the two arms of the inverter bridge are connected to two filter coils 3, each having the inductance L _f . Two non-polarized AC capacitors 4 of equal capacity C _d are connected to the output terminals of the coils, the other two terminals of the capacitors being connected together to one of the DC terminals of the inverter 5, DC + or DC-. Depending on the application in which it is used, the output of inverter 6 in ac (with voltage u ₀ ) can be connected to a single-phase network, single-phase generator, or to a local load. According to this figure, it is indicated that the related control method ensures the transfer of the oscillating power component (p) to the two decoupling capacitors, so that the dc line of the inverter will be traversed only by the average power component (P). at the same time the invented solution will not negatively influence the primary function of the inverter, so that the output quantities u ₀ and i ₀ will not be affected.

RO 130090 Β1

The principle of the control method of the inverter with the decoupling circuit consists in 1 the modification of the control voltages of the two arms of the H-bridge, so that the alternative component of the output power is transferred to the capacitors 4. Considering the case 3 of linear load operation , the instantaneous output power p (t) can be expressed as follows: 5 p (t) = Output ¢ / - Output (2cot - φ) (1) 7 ^P P where: 9

- P, represents the average component of power;

- p, the oscillating component, or a.c., with a frequency equal to twice the frequency of 11 outputs;

According to the method, it is required that the two capacitors with equal values, of the decoupling circuit 13 take over all the alternative component p = p _cdl + p _cd2 , ie:

~ x-ι du ^ _d } „du (·! -,

P ⁼ C _d U _C d \ “! C _d U _cd2 - (2) 17 at the same time, it is required that the output voltage (u ₀ ) is not affected by the decoupling circuit 19, ie:

^u _o ^{= u} cdi ^{- u} cd2 ⁼

Solving the system of equations given by (2) and (3), results the mathematical expression of the voltages at the terminals of the two capacitors, according to the method: 25 u _Cdl2 (t) = + - + - Audiat - φ \ -2υ ² ο sin ² ( ryr) + 4L / ² 0 + - ^ - sin (țz?) ^Cd ^ '2 2] ^ Cd ^[J ° ^{V 7 co} aCd ^V ' ^l comp where: S = U _O I _O - apparent output power; U _o , l ₀ - the effective value of the voltage, respectively 31 of the output current; ω - pulsation of the output voltage; φ - the phase shift between the voltage and the output current; C _d - the capacity of a decoupling capacitor; U _co - an initial voltage per capacitor, which is determined numerically, according to the relation (6) expressed in reported units.

Equation (4) shows the following voltage components on the two capacitors:

- a differential component ± u / 2, which is found at the output of the filter and forms 37 the load voltage u ₀ ,

- a compensation component u ^ p generated in order to perform the proposed decoupling function 39 and which has a common mode character (canceled on the output terminals). 41

For generalization, relation (4) can be written in reported units (u.r.) as follows:

2S 'sin (2ryi - φ) - 2 sin ² (cot) + 4 + ---- sin (^) (5) 45 u' 1 2S ' ^U Cd \, 1

2VQ

RO 130090 Β1 in which for the reported quantities (denoted by the index ‘) the following were taken as basic quantities:

S _n , with c - the nominal apparent power, and (O _n - the pulsation

U b ⁼ U o '^ b ⁼ ^ n' ^ b ^{= n} . a> _n ^v b nominal.

The voltage U ' _c0 can be determined numerically using the following relation:

^ co max <sin 0 <ύ) ί <2π ² (cot) s ^r

---- [sin (2c? r - cp) +>

(6)

It should be noted that, in this analysis, the voltage drop across the L _f filter at the inverter output is neglected at frequencies much lower than the PWM switching frequency, so that the voltage on the two decoupling capacitors is considered approximately equal to the voltage of on the two arms of the inverter (w _cdl2 = Fo). According to the same reasoning, the following approximation is also valid: u _o = u _AB .

Starting from equation (5), fig. 2 shows the main waveforms characteristic of the inverter with the decoupling method according to the invention. The voltages on the two decoupling capacitors, together with the output voltage are shown in fig. 2a, and the main currents in the circuit are represented in fig. 2b. Note that on the dc line the current i _dc contains only the average component (/ _dc ). As a result of the control method according to the invention, the voltages on the two capacitors in the decoupling circuit will contain even order harmonics, the main one being the second order one, but the output voltage not being affected. Also, the currents through the capacitors and the inverter arms will contain higher harmonics, as shown in fig. 2c. in fig. 2d The instantaneous values of the main powers in the circuit were represented over a period of alternating voltage. It is observed that p = p _C di ⁺ Pcd2 such that p _dc = P. Therefore, the decoupling function is fulfilled, as claimed, without the use of additional semiconductor elements.

The method consists, as shown in the block diagram in fig. 1, of a control structure consisting of two subsystems, a subsystem 7 which provides the reference value of the output differential voltage (u _AB ), and another subsystem 8 which generates the compensation component u _comp necessary for the decoupling function. The generation of the two voltages can be done either by direct calculation according to relations (4) or (5), or by using automatic regulators. The signal u _AB multiplied by 1/2 by the amplification block 9, is collected by means of an adder 10 with the signal u, resulting in the control voltage u _A , which divided by means of a division block 11 at the voltage on the line d DC U _dc provides the filling factor D _A for the first arm of the inverter. Similarly, the voltage U _AB , multiplied by 1/2 by the amplification block 9, is subtracted from the voltage u _comp by the adder 12 and results in the control voltage u _B , which divided by a division block 13 at the voltage on the dc line U _dc provides the filling factor D _B for the second arm of the inverter. Next, the resulting filling factors D _A and D _B are transmitted to a PWM pulse generator 14, which supplies 4 control pulses for the transistors T, - T ₄ within the inverter bridge.

FIG. 3 shows relevant experimental results to demonstrate the functionality of the single-phase inverter with the decoupling circuit and the invented method.

RO 130090 Β1

A laboratory stand was used with a single-phase inverter that supplies a 1kW linear load at a voltage of 230V and a frequency of 50Hz. DC supply voltage is 450V and the PWM switching frequency is 10kHz. 3

Disconnect capacitors have a capacity of C _d = 60pF each. The highlighting of the reduction of the alternative component of the i _dc current was made by comparing the results obtained with the 5 decoupling method invented, compared to the classic case of an inverter in the H-bridge without decoupling. Thus, FIG. 3a and fig. 3b illustrates the current i _dc and the voltage u _dc in the two cases. A significant decrease in the oscillating component is observed. Also, the harmonic analysis of the current i _dc , from fig. 3c shows a substantial attenuation of the main alternating component of 9 100Hz. FIG. 3d and fig. 3e shows the voltages on a decoupling capacitor and on the load, respectively the currents through the capacitor, filter and load. The voltage waveform on a decoupling capacitor is observed to be similar to that of FIG. 2a, and the current through the same capacitor contains, in addition to the low frequency harmonic components, a ripple produced by the 13 switching of the transistors. It should be noted that the output voltage u _o does not suffer from visible harmonic distortions, the total voltage distortion factor being about 2%, so the load supply is not affected by circuit I and the decoupling method.

Claims (1)

Claim 1. Method of decoupling the oscillating power for single-phase inverters, which ensures the decoupling of the oscillating component of the instantaneous power of an inverter (1) in the H-bridge, controlled by width-modulated pulses (PWM), which is fed by a line (2) of direct current, the midpoints of the two arms of the inverter bridge being connected to two filter coils (3), and to the output terminals of the coils being connected two non-polarized capacitors (4) intended to disconnect the oscillating component of the instantaneous output power, the other two terminals of the capacitors being connected together to one of the terminals (5) of DC, DC + or DC- of the inverter, and the differential output (6) of the inverter can supply a load, and is characterized in that it consists of the following stages: - two voltages u _AB and u _comp are generated by a control structure consisting of two main subsystems, one subsystem (7) which provides the reference value of the differential output voltage _AB and another subsystem (8) which provides a compensation voltage ^u _comp calculated according to the apparent power S of the load at the inverter output, the pulsation of the output voltage ω, the phase shift angle φ between the voltage and the load current and the capacitance of the disconnect capacitors C _d - multiply the signal u _AB by a factor of 1/2 in an amplification block (9), the resulting signal is summed by means of an adder (10) with the signal u _comp , resulting in a control voltage u _A , which divided by a splitting block (11) at the voltage on the DC line U _dc , provides a signal with the filling factor D _A for the first arm of the inverter, - subtract the signal u _AB multiplied by a factor of 1/2 in the amplification block (9), from the voltage u _comp in an adder (12), resulting in a control voltage ^u b, which is divided by a division block ( 13) at the voltage on the DC line U _dc , provides a signal with the filling factor D _B for the second arm of the inverter, - the transients T _r T ₄ in the H-bridge of the inverter (1) are controlled by the two signals D _A and D _B by means of a PWM pulse generator (14), thus ensuring the decoupling of the oscillating component of the instantaneous power of the inverter ( 1).