CONTROL OF POWER CELLS OF A SPEED CONTROLLER DEPENDING ON THE RECTIFIED VOLTS

Description

Technical field of the invention

[0001] The invention relates to the management of a speed variator in charge of supplying power to an electrical device such as an electric motor for example. State of the art

[0002] According to a power topology, a high voltage is supplied by a variable speed drive by connecting in series a number of low-voltage converters (which are then called power cells or "power cell" (in English). The control of these power cells allows a multi-level voltage to be provided, with each power cell adding a voltage to achieve successive voltage levels.

[0003] According to European standards, low voltage is defined as voltages between 0 and 1000 volts in alternating current (AC) systems and between 0 and 1500 volts in direct current (DC) systems. High voltage is defined as voltages above 1000 volts in AC systems and above 1500 volts in DC systems.

[0004] For example, the speed variator may include N power cells, N being greater than or equal to 2. When the speed variator provides a three-phase supply, it may include 3*N power cells, with N power cells dedicated to each of the three phases.

[0005] In multi-level variable frequency drive (VFD) topologies, only motor voltage measurements may be available to reduce product cost, rather than individual rectified voltages, or DC bus voltages. These rectified voltages correspond to the output voltages of a rectifier bridge (primarily a diode bridge or thyristor bridge) at the input of each of the VFD's power cells.

[0006] In the multi-level architecture described above, the output voltages of each power cell are generated by applying a duty cycle to the rectified voltage of the power cell. The rectified voltage, or DC bus voltage, of each power cell is therefore necessary to calculate the duty cycles to be applied in order to achieve a target output voltage on the motor.

[0007] According to prior art solutions, a fixed rectified voltage common to all power cells is considered, based on a theoretical value of the DC buses in the ideal case.

[0008] However, such a solution introduces an error in the output voltage applied to the motor because, in practice, the rectified voltages of the power cells are different from one another. Indeed, various disturbances are present and affect the rectified voltage levels of each power cell (input voltage of the variable speed drive, transformer, etc.).

[0009] Such an error can be compensated for by measuring motor voltages, which, however, requires accurate motor voltage sensors so that a controller can regulate the voltages generated at the output of the variable speed drive. The voltage reference can therefore consist of two terms: a direct term and a correction from the controller. The larger the error in the direct term, which relies on knowledge of the rectified voltages, the more the controller must work to reject this error. The voltage realization will thus depend on the controller's dynamics and impacts the overall performance of the control system.

[0010] Furthermore, additional sensors can be installed to diagnose the power cells. For example, a temperature probe can be installed on the transformer upstream of the variable speed drive. However, no diagnostics are planned for the input stage of the variable speed drive.

[0011] Therefore, there is a need to improve the control of the output voltage delivered by the variable speed drive to the motor by rejecting various disturbances from the electrical power conversion system (variable transformer gain, dead time during power cell switch switching). Document DE 102014008399 discloses an electronic circuit that does not require voltage measurement for each subunit of the circuit. However, this circuit still requires several sensors.

[0012] The present invention solves the drawbacks mentioned above. Description of the invention

[0013] A first aspect of the invention relates to a method for determining the output voltages of rectifiers as mentioned in claim 1.

[0014] The term "motor voltage" is used below to refer to a voltage across the terminals of the electrical device powered by the speed variator.

[0015] Thus, it is possible to access all rectified voltage values ​​while minimizing the number of measurements taken.

[0016] P is equal to the number of power cells in the speed variator and, at each iteration, a single power cell can be activated, the power cells activated for two different iterations being distinct.

[0017] Such an embodiment allows easy access to each of the rectified voltages of the power cells.

[0018] The method according to the invention thus makes it possible to access the rectified voltages of the power cells, without requiring the addition of additional sensors.

[0019] The process further includes, during a current phase of supplying the electrical device, for each power cell, adapting a control order for the power cell according to the output voltage of the rectifier associated with said power cell.

[0020] This embodiment thus allows for the achievement of a target motor voltage requiring little or no correction. Improved dynamics in achieving the target motor voltage are therefore possible.

[0021] In addition, adapting a control order for a power cell includes determining a duty cycle of the power cell as a function of the output voltage of the rectifier associated with the power cell.

[0022] According to one embodiment, the method may further include analyzing the rectifier output voltages obtained in order to detect a deviation from nominal operation.

[0023] This makes it possible to predict faults in the power stage and to carry out preventive maintenance.

[0024] According to one embodiment, the process can be repeated at several distinct times, and the analysis of the rectifier output voltages can include determining trends in the evolution of the rectifier output voltages of the power cells, and comparing the trends with nominal operation.

[0025] This makes it possible to predict faults in the power stage and to carry out preventive maintenance.

[0026] According to one embodiment, the process can be initiated following the shutdown of the electrical device.

[0027] Thus, the process does not interrupt the operation of the electrical device.

[0028] In addition, the process can be initiated automatically upon detection of the electrical device stopping.

[0029] No operator intervention is required. In addition, the process can be initiated automatically at a given frequency when the electrical device is switched off.

[0030] Alternatively, the process can be initiated manually following the shutdown of the electrical device.

[0031] An operator can therefore carry out a check of the operation of the power cells.

[0032] A second aspect of the invention relates to a computer program executable by a processor and comprising instructions to, when executed by the processor, implement the steps of a process according to the first aspect of the invention.

[0033] A third aspect of the invention relates to a control device for a speed variator as stated in claim 9.

[0034] A fourth aspect of the invention relates to a speed variator as stated in claim 10. Brief description of the figures

[0035] By way of example only, embodiments of the invention will be described with reference to the drawings, among which: [ Fig. 1 ] - there figure 1 illustrates a system for controlling the power supply of an electrical device according to an embodiment of the invention; [ Fig. 2 ] - there figure 2 is a diagram illustrating the steps of a process according to an embodiment of the invention; [ Fig. 3 ] - there figure 3 illustrates the structure of a control device according to one embodiment of the invention. Detailed description

[0036] There figure 1This presents a power supply system for an electrical device, such as an electric motor 100 (for example, an induction motor), powered by a three-phase variable power supply. Such a motor is given by way of illustration, but the invention is not limited to this single example; the invention is applicable to any electrical device powered by a variable speed drive comprising several power cells.

[0037] The speed variator includes a transformer 111 receiving a variable three-phase supply from the mains 110. The transformer 111 can be a multi-winding transformer capable of delivering three-phase voltages to several power cells described below.

[0038] The speed variator 102 according to the invention may include a power stage comprising one or more low-voltage power cells 101. In the example illustrated on the figure 1The motor receives a three-phase supply and the 102 speed variator includes 3*N power cells, with N power cells dedicated to each phase, N being greater than or equal to 2.

[0039] With reference to the figure 1A system with 3*N power cells is shown. However, the invention applies equally to a system with 3*(N+1) power cells, with one power cell reserved per phase in case of failure of one of the active power cells. Such an example is given for illustrative purposes only. The example of a three-phase power supply is considered, but the invention also applies to a system with N or N+1 power cells. The invention also applies to a system with N+2 (or more than N+2) power cells (or 3*(N+2) for a three-phase supply), with two replacement power cells per phase. Furthermore, the invention is applicable to a number of phases other than three. In addition, the number of power cells may differ per phase, in which case each phase comprises a number Ni of power cells, i being a phase index.In the following discussion, the case of a three-phase power supply with the same number of power cells N for each phase is considered for illustrative purposes only.

[0040] Each power cell 101 receives as input the three-phase power supply from the secondary of the transformer 111 and may include a rectifier (not shown on the figure 1 At the input, the rectifier is capable of rectifying the incoming three-phase power supply to provide a direct current (DC) voltage. The rectified DC voltage obtained for each power cell 101 is also called the DC bus voltage, or bus voltage. The rectifier may include a diode bridge, a thyristor bridge, or any other system known to rectify a voltage.

[0041] At the output of the rectifier, each power cell 101 may include a capacitor suitable for storing electrical energy, as well as a module for generating a pulse width modulation (PWM) signal (for " Pulse Width Modulation » (in English). Such a generation module may include an H-bridge comprising four switches controlled in pairs. A power electronics system using such a chopped voltage principle applies to the motor, per phase, a voltage that is proportional to one or more rectified voltages. On average, the applied proportion corresponds to the ratio between the target output voltage of the power cell and the reference rectified voltage (defined below). The operation of an H-bridge is well known and will not be described further in this application.

[0042] The switches in the H-bridge can be IGBT type transistors (for " Insulated Gate Bipolar Transistor" which have the advantage of being able to be switched quickly.

[0043] The switches of a power cell 101 are controlled by a control cell 103 of the power cell 101.

[0044] The system further includes a control device 120 capable of controlling the operation of the power cells 101 of the variable speed drive 102 in order to control the power supply to the motor 100. For this purpose, the control device 120 can control the control cells 103 of the power cells 101. The control device 120 can also control switches 104 for connecting a subset of the N power cells for each phase in series. Alternatively, these switches are controlled by the control device 120 via the control cells 103.

[0045] The power cells 101 can receive control signals from the control device 120 from which the power cells 101 can control the switching of the H-bridge switches.

[0046] The three-phase voltages, or motor voltages, supplying motor 100 are thus obtained by summing the PWM output voltages of the power cells 101, indicated as V cell on the figure 1 , for which switches 104 are open (called " active cells » below). Switches 104 allow you to « bypass » ( bypass (in English) the power cells 101 without controlling the switches of the power cell H-bridge. Such switches 104 are, however, optional in that the power cells 101 can be deactivated by controlling their H-bridge so as to have a zero duty cycle, and therefore a zero output voltage of the power cell.

[0047] The system according to the invention further comprises a means 130 for measuring the three-phase voltages supplying the motor. The measuring means 130 is capable of transmitting the measured motor voltage(s) to the control device 120.

[0048] The control device 120 according to the invention is configured to control the various power cells 101, as well as the bypass switches, directly or indirectly (directly or via the control cells 103). In particular, according to the invention, the control device 120 is capable of controlling the power cells 101 or the bypass switches 104 so that certain cells do not supply voltage, and thus the motor voltage is supplied by only one power cell, for example. Thus, in an embodiment in which only one power cell is activated at a time, a measured motor voltage provides access to the output voltage of the selected power cell, as detailed later, and it thus provides access to the DC bus voltage from the duty cycle applied to the power cell.

[0049] There figure 2is a diagram illustrating the steps of a process according to an embodiment of the invention.

[0050] At step 200, the process according to the invention is implemented. Such implementation may occur following the shutdown of the motor 100. For example, following the shutdown of the motor 100, the process is triggered automatically, or manually, via a human-machine interface of the control device 120.

[0051] At a step 201, the control device 120 controls the switches 104 or the H-bridges of the different cells, so as to activate a subset of the power cells of the variable speed drive 102, for example only the first cell 101 of the first phase. The operation of embodiments in which several power cells are activated simultaneously will be described later. figure 2being described within the framework of an embodiment based on successive activations of power cells (only one power cell activated at each iteration)

[0052] In parallel, at a step 202, the control device 120 determines a voltage value to be applied to the speed variator 102. This voltage to be applied can be direct, or calculated by a current regulator for example, allowing control of the output current of the speed variator passing through the motor 100.

[0053] At stage 203, a cyclic report, or " duty cycle » , to be applied to the activated power cell 101 1,1 is determined by the control device 120 in order to achieve a target output voltage V 0 to be applied to the motor 100. For the other inactive cells, either the bypass switches 104 are closed, or the duty cycles are zero.

[0054] The speed variator 102 is powered in step 204 and the power cells 101 are controlled in accordance with steps 202 and 203.

[0055] The motor voltage is then measured at a step 205, for example by a sensor, and then received by the control device 120, the motor voltage being supplied by the first cell 101 1,1.

[0056] The control device 120 determines at step 206 whether all the power cells 101 have been successively activated or not. If so, the process proceeds to step 207. More generally, a predefined number of iterations P is defined, P being greater than or equal to 2. Thus, an iteration index can be compared to P at step 207. If the iteration index is equal to P, the process proceeds to step 207.

[0057] Otherwise, the process returns to step 201, activating the next power cell (for example, power cell 101 1,2 after power cell 101 1,1), or subsequent power cells, in the general case which will be described later, and deactivating all other power cells 101. In an alternative embodiment, the process returns to step 201 but also to step 202. In this alternative embodiment, steps 202 and 203 are implemented for each iteration of the process, in a differentiated manner.

[0058] From the motor voltages measured in successive steps 205, the rectified voltages of the power cells are deduced in a step 206 by the control device 120 according to the invention.

[0059] In order to determine the rectified voltage of a power cell, the control device 120 takes into account the measured motor voltage and the duty cycle that has been applied.

[0060] It should be noted that, in the embodiment in which the process is iterated 3*N times (in the case where the speed variator includes N power cells per phase among three phases) by activating a single power cell each time, the rectified voltage of the power cell activated during a given iteration can be determined at the end of the given iteration (and not at the end of all the iterations).

[0061] After determining the rectified voltages of the power cells 101 of the speed variator 102, the process proceeds to step 208.

[0062] In step 208, the rectified voltages of the power cells of the variable speed drive are stored in a memory of the control device 120. The rectified voltages can be stored as replacements for previous rectified voltage values. Alternatively, the rectified voltages are stored in association with the implementation date of the iteration of the process according to the invention, thus enabling monitoring of the evolution of the rectified voltages for each of the power cells 101 of the variable speed drive 102.

[0063] In the optional step 209, an analysis of the rectified voltages is implemented. Such an analysis may include a comparison with nominal rectified voltages and the triggering of an alert in the event of a deviation from one of the nominal rectified voltages. If the rectified voltages are stored in association with their respective dates, a trend in the evolution of the rectified voltage can be determined for each power cell 101, and an alert can be generated based on this trend.

[0064] Generating an alert allows for the implementation of preventive maintenance.

[0065] Regardless of steps 208 and 209, the speed variator switches to "activated" or "ready" mode at step 210, allowing motor 100 to start at step 211.

[0066] Following steps 209 and 210, the control device 120, upon receiving a motor command, determines a target motor voltage and controls the speed variator 102 according to the target motor voltage and according to the rectified voltages previously determined in step 211. For this purpose, the control device 120 determines respective duty cycles for the power cells 101. The achievement of the target motor voltage is then more precise in that the rectified voltages of the power cells are known and can be updated regularly (for example, at each stop of the electric motor 100).

[0067] According to some embodiments, such a speed variator further includes a compensator correcting the input commands of the control device 120 according to the voltages measured on the electric motor 100. In this case, the role of such a compensator is reduced according to the invention, which makes it possible to improve the dynamics of the realization of the output voltages applied to the motor 100.

[0068] The process can then be repeated at a later time, for example following a new stop of the electric motor 100 at a step 200.

[0069] The process can alternatively be repeated at a given frequency, for example, weekly, following the shutdown of the electric motor. In this case, a minimum interval of one week separates the implementations of two iterations of the process.

[0070] With reference to the figure 2A particular embodiment involving the activation of a single power cell at each iteration has been described. Such a method is simple to implement, and the rectified voltage of each power cell is readily accessible once the motor voltages have been measured. More generally, the invention can provide, at each iteration, for the activation in step 201 of a subset of at least one power cell, the realization of an output voltage or output voltages by the subset in step 204, and the measurement of one or two motor voltages in step 205. Indeed, it is not necessary to measure more than two motor voltages since a third motor voltage can be deduced from the first and second motor voltages (see voltages U12, U23, and U31 described below).

[0071] The power cells to be activated at each iteration and the motor voltages to be measured at each iteration can be determined from a predefined matrix. This matrix is ​​predefined to ensure that, after all iterations, a sufficient number of equations have been obtained to determine the value of each unknown in the system of equations (for example, the 3*N rectified voltages of the power cells when the variable speed drive comprises 3*N power cells). To this end, the predefined matrix has a rank equal to the number of unknowns (i.e., 3*N).

[0072] Let i be the phase index (ranging from 1 to 3), and k the power cell index within a phase (ranging from 1 to N in a power stage with 3*N power cells), a power cell of phase i and index k: is activated with a duty cycle r (between -1 and 1); has a rectified voltage VB(i,k), which is an unknown of the system; provides a potential difference rx VB(i,k)

[0073] Column vectors VBi with N components can be defined, each component with index k being the value of the rectified voltage VB(i,k).

[0074] A combination of cell activation commands from an arm of the power stage dedicated to phase i can be represented by the line vector Ki with N values ​​(0: power cell not active, 1: power cell active).

[0075] The potential difference Vi generated by the arm with index i is equal to the sum of the potential differences of all the cells in the arm, that is: Vi = Ki x VBi The measuring means 130 described above can be adapted to provide three phase-to-phase line voltages U12 = V1-V2, U23 = V2-V3, and U31 = V3-V1. Alternatively, the measuring means 130 is arranged to measure the voltages V1, V2, and V3 directly. The three measurements at a time t corresponding to an iteration of the process (t can be a date or an iteration index of the process), in particular of step 205, can be associated in a column vector M_t equal to [U12; U23; U31], i.e.: M_t = KALL_t x VBALL, where K0 is a zero vector of the same dimension as K1, K2, or K3, and KALL_t represents a 3x3, 3*N column matrix composed as follows: [K1 - K2 K0; K0 K2 - K3; -K1 K0 K3]; and VBALL represents a column vector of length 3N composed as follows: [ VB1 ; ​​VB2 ; VB3 ].

[0076] The combinations K1 to K3 are chosen according to the invention so that a matrix KALL_t ALL, with 3*P rows and 3*N columns, constructed for the times (or iterations) t1, t2, up to tP by "stacking" or aggregation of the matrices KALL_t u for u varying between 1 to P, KALL_tALL is a matrix of rank 3*N, so as to be able to determine the 3*N unknowns of the system (the rectified voltages of the power cells).

[0077] The method of implementation described with reference to the figure 2 is a special case of such a general embodiment.

[0078] This particular embodiment involves the activation of a single cell at each time tu, u being an index between 1 and 3*N.

[0079] When activating the index cell "u- INT[(u-1) / N] *N" of the "1 + INT[(u-1) / N]" arm of the power stage, where INT represents the integer part function, we obtain, with a case in which N=3 (given as an example, without restricting the scope of the invention): At time t1, K1 = [1 0 0]; K2 = [0 0 0]; K3 = [0 0 0]. The matrix KALL_t1 is [1 0 0 0 0 0 0 0 0 ; 0 0 0 0 0 0 0 0 0 ; -1 0 0 0 0 0 0 0 0]; at time t2, K1 = [0 1 0]; K2 = [0 0 0]; K3 = [0 0 0]. The matrix KALL_t2 is [0 1 0 0 0 0 0 0 0 ; 0 0 0 0 0 0 0 0 0 ; 0 -1 0 0 0 0 0 0 0].

[0080] By construction of each KALL_tu matrix, the first row consists of 0s except for the u-th element which is worth 1. It is understood that after 3x3 times according to the rule explained above, the first 9 rows taken together form the identity matrix, which guarantees rank 9 (i.e. the number of unknowns in the system).

[0081] There figure 3illustrates the structure of a control device 120 according to one embodiment of the invention.

[0082] The control device 120 includes a processor 300 configured to communicate unidirectionally or bidirectionally, via one or more buses, with a memory 301 such as a " Random Access Memory » , RAM, or a memory type " Read Only Memory » , ROM, or any other type of memory (Flash, EEPROM, etc.).

[0083] The memory 301 is capable of storing, permanently or temporarily, at least some of the data used and / or generated during the implementation of the process according to the invention. In particular, the memory 301 is capable of storing the rectified voltage values ​​of the power cells, optionally in association with their respective dates.

[0084] The 300 processor is capable of executing instructions for implementing the steps of the process according to the invention, illustrated with reference to the figure 2 .

[0085] The control device 120 may further include an input interface 302 and an output interface 303 in order to communicate with the other entities of the system according to the invention.

[0086] In particular, the input interface 302 is capable of receiving speed commands and the processor 300 is capable of determining the target motor voltages (one target motor voltage per phase) from the speed commands.

[0087] The output interface 303 is capable of providing control commands to the power cells 101 so that they perform duty cycles, particularly via their H-bridges. For this purpose, the control commands can include switching commands for IGBT-type switches, for example, partially (only the commands for the two uppermost IGBT switches for each power cell) or fully (all four IGBT switches). These switching commands can be calculated by the control device 120 by comparing the reference voltages with triangular signals corresponding to each power cell. This technique is well known and will not be described in further detail.

[0088] Although the present invention has been described above with reference to particular embodiments, the invention is in no way limited to the forms described. The invention is limited only by what is defined in the claims, and other embodiments than those described above may fall within the scope of the claims.

Claims

1. A method for determining rectified voltages of power cells of a variable-speed drive (102) in charge of powering an electrical device (100), the variable-speed drive comprising Ni low-voltage power cells (101) connected in series for each phase among several phases, N being greater than or equal to 2, and i being a phase index, characterised in that the method comprises the following operations of: a) repeating the next P iterations, P being equal to the number of power cells of the variable-speed drive: i) activating (201) at each iteration of one single cell of a phase and deactivating the other power cells of said phase of the variable-speed drive, wherein said activated cell is selected according to predefined activation commands dependent on an iteration index; the power cells activated for two different iterations being distinct; ii) determining (203) a duty cycle of said cell activated to control an output voltage by pulse width modulation of said cell activated in order to achieve an output target voltage to be applied to the electrical device (100); iii) receiving (205) at least one output voltage of the variable-speed drive (102) across the electrical device (100) for at least one phase, and wherein the method further comprises, at the end of the P iterations: b) determining (206), from the measured output voltages of the variable-speed drive, rectified voltage values at the output of respective rectifying stages of the power cells of the variable-speed drive; c) storing (208) the determined rectified voltage values, respectively associated with the power cells of the variable-speed drive, d) for each power cell, adapting a control command of the power cell comprising determining a duty cycle of the power cell as a function of the rectified voltage associated with this power cell, during a current phase of supplying the electrical device (100), wherein the integer P and the activation commands are predefined in the form of a matrix representing the activation commands and with a rank equal to the number of power cells of the variable-speed drive and wherein the variable-speed drive comprises N power cells for each phase among three phases and, for each iteration of index t, t being between 1 and P: M_t = KALL_t × VBALL; wherein KALL_t represents a 3-line and 3*N column matrix comprising [K1 - K2 K0; K0 K2 -K3; -K1 K0 K3], wherein K1, K2 and K3 are the respective activation commands of the three phases and are N-size vectors taking binary values, a first value corresponding to an activation and a second value corresponding to a deactivation, wherein K0 is a N-size zero vector, wherein VBALL is a column vector with a length 3N composed as follows [VB1; VB2; VB3], wherein VB1, VB2 and VB3 are N-size vectors of the rectified voltages of the power cells of the respective three phases according to the duty cycle determined in step ii); wherein M_t is a vector of three voltages measured across the electrical device for a given iteration, the measured voltages corresponding to the three output voltages of the variable-speed drive.

2. The method according to claim 1, wherein adapting a control command of a power cell (101) comprises determining a duty cycle of the power cell (101) as a function of the rectified voltage associated with the power cell.

3. The method according to one of the preceding claims, further comprising analysing (209) the rectified voltages obtained with a view to detecting a deviation with respect to nominal operation.

4. The method according to one of claims 1 to 3, wherein the method is repeated at several distinct instants, and wherein analysing (209) the rectified voltages comprises determining changing trends of the rectified voltages of the power cells, and comparing the trends with nominal operation.

5. The method according to one of the preceding claims, the method being initiated subsequent to stopping the electrical device (100).

6. The method according to claim 5, wherein the method is automatically initiated upon detecting stopping of the electrical device (100).

7. The method according to claim 5, wherein the method is manually initiated subsequent to stopping the electrical device (100).

8. A computer program executable by a processor (300) and comprising instructions for, when executed by the processor, implementing the steps of a method according to any one of claims 1 to 7.

9. A device for controlling a variable-speed drive (102) in charge of powering an electrical device (100), the variable-speed drive comprising Ni low-voltage power cells (101) connected in series for each phase among several phases, N being greater than or equal to 2 and i being a phase index, characterised in that the control device comprises: a processor (300) capable of controlling, by an output interface, the repetition of the next P iterations, P being equal to the number of power cells of the variable-speed drive: i) activating (201) at each iteration of one single cell of a phase and deactivating the other power cells of said phase of the variable-speed drive, wherein said one activated cell is selected according to predefined activation commands dependent on an iteration index; the cells activated for two different iterations being distinct; ii) determining a duty cycle of said cell activated to control an output voltage by pulse width modulation of said cell activated in order to achieve an output target voltage to be applied to the electrical device (100); iii) receiving, via an input interface (302), at least two output voltages of the variable-speed drive (102) across the electrical device (100) for two phases; the processor (300) being further capable of determining, at the end of the P iterations, from the measured output voltages of the variable-speed drive, rectified voltage values at the output of respective rectifying stages of the power cells of the variable-speed drive, a memory (301) storing the determined rectified voltage values, respectively associated with the power cells of the variable-speed drive; the processor being configured to adapt for each power cell, a control command of the power cell by determining a duty cycle of the power cell as a function of the rectified voltage associated with this power cell, wherein the integer P and the activation commands are predefined in the form of a matrix representing the activation commands and with a rank equal to the number of power cells of the variable-speed drive and wherein the variable-speed drive comprises N power cells for each phase among three phases and, for each iteration with an index t, t being between 1 and P: M_t = KALL_t × VBALL; wherein KALL_t represents a 3-line and 3*N column matrix comprising [K1 - K2 K0; K0 K2 -K3; -K1 K0 K3], wherein K1, K2 and K3 are the respective activation commands of the three phases and are N-size vectors taking binary values, a first value corresponding to an activation and a second value corresponding to a deactivation, wherein KO is a N-size zero vector, wherein VBALL is a column vector with a length 3N composed as follows [VB 1; VB2; VB3], wherein VB1, VB2 and VB3 are N-size vectors of the rectified voltages of the power cells of the respective three phases according to the duty cycle determined in step ii); wherein M_t is a vector of three voltages measured across the electrical device for a given iteration, the measured motor voltages corresponding to the three output voltages of the variable-speed drive.

10. A variable-speed drive in charge of three-phase power supply of an electrical device, the variable-speed drive comprising Ni low-voltage power cells connected in series for each phase among several phases, N being greater than or equal to 2 and i being a phase index, and a control device according to claim 9.