AC-AC converter comprising a matrix array of bidirectional switches of programmable configuration

Abstract

An AC-AC matrix converter that converts an input multiphase periodic voltage includes N input voltages that are out of phase into an output multiphase periodic voltage comprising N output voltages that are out of phase, the converter comprising a square matrix array comprising N.sup.2 switches. The converter comprises command and control electronics that periodically perform the following two functions: storing N! voltage summations, each voltage summation corresponding to one switch configuration, each switch configuration relating one and only one out-of-phase input voltage and one and only one out-of-phase reference voltage, each voltage summation being the summation of the N differences in absolute value between one and only one out-of-phase input voltage and one and only one out-of-phase reference voltage; switching the matrix array of switches to apply the configuration corresponding to the lowest voltage summation.

Claims

1. An AC-AC matrix converter that converts an input multiphase periodic voltage comprising N input voltages into an output multiphase periodic voltage comprising N output voltages, said converter comprising a square matrix array comprising N.sup.2 switches, wherein the converter comprises command and control electronics that periodically perform at least the two following functions: storing N! voltage summations, each voltage summation corresponding to one switch configuration, each switch configuration relating N times one and only one input voltage and one and only one reference voltage, a reference voltage corresponding to an expected output voltage, each voltage summation being the summation of the N differences in absolute value between one and only one input voltage and one and only one reference voltage switching the matrix array of switches to apply the switch configuration corresponding to the lowest voltage summation.

2. The AC-AC matrix converter as claimed in claim 1, wherein the converter comprises a sample-and-hold module that periodically stores, at the sampling frequency, the N! voltage summations, said sampling frequency being an order of magnitude higher than the frequency of the input periodic voltage and an order of magnitude higher than the frequency of the output periodic voltage.

3. The AC-AC matrix converter as claimed in claim 2, comprising determining means that periodically determine, at the sampling frequency, the lowest voltage summation among the N! voltage summations.

4. The AC-AC matrix converter as claimed in claim 2, wherein the converter comprises a first assembly for filtering the input multiphase voltage, the matrix array of switches, a second assembly for filtering the output multiphase voltage, an electronic control loop and the command and control electronics, the command and control electronics controlling the electronic control loop and the electronic control loop controlling the switching of the matrix array of switches.

5. The AC-AC matrix converter as claimed in claim 4, wherein the electronic control loop is configured to command the switching of the matrix array of switches in such a way as to command to open and to close each switch defined as being closed in said switch configuration corresponding to the lowest voltage summation, at a chopping frequency and with a duty cycle that are defined by the control loop on the basis of the measurements of the output phase currents, and to keep open each switch defined as being open in said switch configuration corresponding to the lowest voltage summation.

6. The AC-AC matrix converter as claimed in claim 1, wherein the number N being equal to 3, the input and output voltages are three-phase, the matrix array of switches comprising 9 switches and the number of switch configurations being equal to 6.

7. The AC-AC matrix converter as claimed in claim 1, wherein the number N being equal to 4, the input and output voltages are four-phase, the matrix array of switches comprising 16 switches and the number of switch configurations being equal to 24.

8. The AC-AC matrix converter as claimed in claim 1, wherein the frequencies of the input periodic voltage and of the output periodic voltage are comprised between 0.1 Hz and 1000 Hz.

9. The AC-AC matrix converter as claimed in claim 1, wherein the powers transmitted by said converter are comprised between 1 kW and 1000 kW.

10. The AC-AC matrix converter as claimed in claim 1, wherein the converter is bidirectional, i.e. the input multiphase periodic voltage is generated by a voltage source and the output multiphase periodic voltage powers a load or vice versa.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Other features, details and advantages of the invention will become apparent upon reading the description, which is given with reference to the appended drawings, which are given by way of example and in which, respectively:

(2) FIG. 1 is an illustration of a first prior-art AC-AC converter;

(3) FIG. 2 is an illustration of a second prior-art AC-AC converter;

(4) FIG. 3 is an illustration of a three-phase AC-AC matrix converter;

(5) FIG. 4 is an illustration of a four-phase AC-AC matrix converter;

(6) FIG. 5 shows various types of switches of a matrix converter;

(7) FIG. 6 shows various perturbations of a three-phase signal;

(8) FIG. 7 shows a phase-locked loop (PLL) used in a prior-art matrix converter;

(9) FIG. 8 shows the architecture of a matrix array of switches of a converter according to the invention;

(10) FIG. 9 shows the architecture of a matrix converter according to the invention in the case of three-phase input and output voltages;

(11) FIG. 10 shows the architecture of the command electronics of a matrix converter according to the invention;

(12) FIG. 11 shows the six switch configurations in the case of three-phase signals;

(13) FIG. 12 shows the architecture of the command electronics of a matrix converter according to the invention in the case of source-to-load operation;

(14) FIG. 13 shows the architecture of the command electronics of a matrix converter according to the invention in the case of load-to-source operation.

DETAILED DESCRIPTION

(15) The AC-AC matrix converter according to the invention is applicable to any type of multiphase voltage. However, as has been seen, the number of possible switch configurations increases as the factorial of the number of phases. Thus, below, for the sake of clarity, FIGS. 9 to 13 relate to three-phase voltages, thus limiting the number of switch configurations to 6. For a person skilled in the art, the generalization to voltages comprising more than three phases will present no technical difficulty, the number of configurations to be taken into account simply being higher, the architecture of the converter on the whole remaining the same.

(16) FIG. 9 shows the architecture of a matrix converter 100 according to the invention in the case of three-phase input 200 and output 300 voltages. Below, the three phase voltages of the input voltage will be denoted A, B and C, and the three phase voltages of the output voltage will be denoted U, V and W. Furthermore, the values of the three reference voltages at a time t corresponding to a sampling time will be denoted Vu, Vv and Vw, and the values of the three output voltages at the same sampling time t will be denoted Vu, Vv and Vw.

(17) This matrix converter 100 comprises the following elements:

(18) a first filtering assembly 101;

(19) a matrix array 102 of switches;

(20) a second filtering assembly 103;

(21) command and control electronics 110;

(22) an electronic control loop 104;

(23) electronics 105 for measuring currents.

(24) The first filtering assembly 101 is placed between the voltage source 200 and the matrix array 102 switches. The second filtering assembly 103 is placed between the matrix array 102 switches and the load 300. Each filtering assembly comprises three identical filtering devices placed on the three input channels and on the three output channels of the matrix converter. The matrix array 102 of switches comprises three rows of three switches (the latter are not shown in FIG. 9).

(25) The command and control electronics 110 controls, i.e. commands, the electronic control loop 104 or pulse width modulation (PWM), which commands the switches of the matrix array 102 to open and to close at the frequency of the PWM, which is called the chopping frequency.

(26) The electronics of the PWM are also controlled, i.e. commanded, in current-mode via the measurements of current of the electronics 105, i.e. of the phase currents output from the matrix array 102 and as measured by the electronics 105. In other words, the control loop 104 commands the switches of the matrix array 102 to open and to close at the frequency of the PWM on the basis of measurements of the measured phase currents output from the matrix array 102.

(27) The duration of application of the PWM is proportional to the desired current.

(28) In other words, the duty cycle of the PWM is proportional to the desired current.

(29) The function of the control loop is to command the switches of the matrix array 102 of switches so as to regulate the output RMS voltage or the output RMS current to a set point RMS voltage or a set point RMS current, respectively. The command and control electronics 110 comprises the following elements:

(30) three means 111, 112 and 113 for measuring the three input voltages Va, Vb and Vc at a time t corresponding to a sampling time;

(31) three means 114, 115 and 116 for computing three setpoints Vu, Vv and Vw corresponding to the expected values of the three output voltages at the same sampling time t;

(32) six means 121 to 126 for computing voltage summations, each voltage summation corresponding to one switch configuration. These various configurations are shown in FIG. 11. They are denoted C1 to C6. More precisely, at each sampling time:
|Va−Vu|+|Vb−Vv|+|Vc−Vw|  Configuration C1
|Va−Vu|+|Vb−Vw|+|Vc−Vv|  Configuration C2
|Va−Vv|+|Vb−Vu|+|Vc−Vw|  Configuration C3
|Va−Vv|+|Vb−Vw|+|Vc−Vu|  Configuration C4
|Va−Vw|+|Vb−Vu|+|Vc−Vv|  Configuration C5
|Va−Vw|+|Vb−Vv|+|Vc−Vu|  Configuration C6

(33) the sample-and-hold module 140, the function of which is to periodically keep constant, i.e. to periodically store, at the sampling frequency, the values of the voltage summations. The values of the summations of the voltages, i.e. of the configurations, correspond to the results of the latter computations, which are performed by the computing means 121 to 126 and received by the sample-and-hold module.

(34) control means 130, which operate at the clock frequency of the sampling, and control, i.e. command, the various electronic control means on each change of sampling time t, the sampling frequency being an order of magnitude higher than the frequency of the input periodic voltage and an order of magnitude higher than the frequency of the output periodic voltage;

(35) comparing means 150, which compare the values of the six voltage summations computed by the computing means 121 to 126;

(36) means 160 for determining the lowest voltage sum;

(37) optionally, means 170 for commanding switching of the switches so as to apply the configuration corresponding to the lowest voltage summation;

(38) an optional power-controlling interface 180 that implements the preceding commands.

(39) The control means 130 command the means 150, 160 so that the comparing means 150 periodically compare, at the sampling frequency, the values of the voltage summations, and so that the determining means 160 periodically determine, at the sampling frequency, the lowest voltage summation corresponding to the optimal switch configuration.

(40) In the embodiments in FIGS. 10 and 12, the command and control electronics 110 are devoid of control means 170 and 180.

(41) The control loop 104 receives the lowest voltage determined by the determining means.

(42) In a switch configuration corresponding to the lowest voltage sum, each switch of the matrix array 102 is defined as being, either closed, or open. The control loop 104 is configured to command the switching of the matrix array 102 of switches in such a way as to command to open and to close each switch of the matrix array 102 that is defined as being closed in said switch configuration corresponding to the lowest voltage summation, at a chopping frequency and with a duty cycle that are defined by the control loop on the basis of the measurements of the output phase currents, and to keep open each switch defined as being open in said switch configuration corresponding to the lowest voltage summation. On each change of sampling time, the voltage measurements, the computations and the selection of the best configuration are reiterated, the best configuration possibly being identical to that selected at the preceding time or another configuration.

(43) The frequencies of the periodic input voltage and of the periodic output voltage are preferably comprised between 50 Hz and 1000 Hz, though the converter according to the invention is able to work at lower or higher frequencies.

(44) The powers transmitted by said converter are comprised between 1 kW and 1000 kW, this range of values being nonlimiting.

(45) On account of the operating frequency and power values, the electronic components to be employed present no particular difficulties.

(46) The converter according to the invention may be bidirectional, i.e. the input multiphase periodic voltage is generated by a voltage source and the output multiphase periodic voltage powers a load or vice versa.

(47) FIGS. 12 and 13 illustrate this reversibility. In FIG. 12, the converter is placed between an input source S and an output load C. In FIG. 13, the converter is placed between an input load C and an output source S.

(48) In these two configurations, the structure of the converter is identical. It comprises, as described above, a first filtering assembly 101, a matrix array 102 of switches, a second filtering assembly 103, command and control electronics 110, an electronic control loop 104 and electronics 105 for measuring currents.

(49) In the example detailed above, the voltages were three-phase voltages. When the voltages are four-phase voltages, the architecture of the converter comprises the same elements as that of FIG. 9. The number of switch configurations to compare is simply higher. It increases to 24 configurations. The number of configurations would increase to 120 in the case of voltages comprising five phases and so on.

(50) As one variant, the converter according to the invention is devoid of electronic control loop 104 and optionally of electronics 105 for measuring currents. The switches are then commanded on the basis of the voltage values and not on the basis of measurements of currents output from the matrix array. The command and control electronics 110 then comprise the elements 170 and 180, which are configured to apply the switch configuration corresponding to the lowest voltage summation determined by the means 160.