Device for controlling a polyphase inverter

Abstract

The device according to the invention controls a polyphase inverter (10, 14, 17) intended for powering from a DC current source (CC) a polyphase rotating electric machine (1). The device is of the type of those generating commutation functions driving commutation elements (9, 13) of the inverter in such a way as to obtain a reduction of the losses in the commutation elements and a decrease of an effective current in a decoupling capacitor (16) of the source (CC). According to the invention, this reduction and this decrease are obtained by means of a set of control strategies (21, 24) determining the commutation functions by using additional degrees of freedom of the polyphase machine (1) with respect to a three-phase reference machine. The polyphase machine comprises first and second phase windings forming a first three-phase system (2, 3, 4) and a second three-phase system (5, 6, 7) with distinct neutral points (11, 15) offset angularly by a predetermined angle of offset (). The first and second phase windings are linked respectively to three first and three second power arms (8, 12) formed by the commutation elements.

Claims

1. A device for controlling a polyphase inverter for supplying power from a direct current source to a double three-phase rotating electrical machine, said device comprising: means for generating switching signals controlling switching elements to obtain a reduction of losses in said switching elements and a reduction in an rms current in a decoupling capacitor of said direct current source, said double three-phase rotating electrical machine comprising three first phase windings and three second phase windings forming a first three-phase system and a second three-phase system with separate neutral points offset angularly by a predetermined offset angle, and said first and second phase windings being connected to three first and three second power arms, respectively, formed by said switching elements of said polyphase inverter; means for storing a set of at least two control strategies determining said switching signals; means for acquiring a rotation speed and a power factor of said machine; and means for selecting a current strategy from said set of control strategies as a function of said rotation speed and said power factor.

2. The device as claimed in claim 1 for controlling a polyphase inverter, wherein said set of control strategies includes a strategy comprising applying to said first three-phase system a first centered vectorial pulse width modulation offset by a common delay from a second centered vectorial pulse width modulation with the same period applied to said second three-phase system.

3. The device as claimed in claim 1 for controlling a polyphase inverter, wherein said set of control strategies includes a strategy comprising in applying to said first three-phase system a first vectorial pulse width modulation offset by a common delay from a second vectorial pulse width modulation with the same period applied to said second three-phase system and blocking one of said three first arms and/or one of said three second arms, said common delay being substantially equal to 25% of said period.

4. The device as claimed in claim 3 for controlling a polyphase inverter wherein said strategy is applied if: a rotation speed of said machine is less than a first predetermined speed representative of an end of starting of said machine, or said rotation speed is between said first predetermined speed and a second predetermined speed representative of an end of functioning at constant torque of said machine, greater than said first predetermined speed, and a power factor of said machine is less than or equal to a predetermined coefficient substantially equal to 0.7.

5. The device as claimed in claim 1 for controlling a polyphase inverter, wherein said set of control strategies includes a strategy comprising applying to said first three-phase system a first generalized discontinuous pulse width modulation offset by a common delay from a second generalized discontinuous pulse width modulation with the same period applied to said second three-phase system, said common delay being substantially equal to 30% of said period.

6. The device as claimed in claim 5 for controlling a polyphase inverter wherein said strategy is applied if: a rotation speed of said machine is between a first predetermined speed representative of an end of starting of said machine and a second predetermined speed, representative of an end of functioning of said machine at constant torque, greater than said first predetermined speed, and a power factor of said machine is greater than a predetermined coefficient substantially equal to 0.7, or said rotation speed is between said second predetermined speed and a third predetermined speed, representative of operation of said machine at constant power greater than said second predetermined speed.

7. The device as claimed in claim 1 for controlling a polyphase inverter, wherein said set of control strategies includes a strategy comprising: applying to said first three-phase system a first centered vectorial pulse width modulation and to said second three-phase system a second centered vectorial pulse width modulation having the same period and the same time origin; blocking one of said three first arms and one of said three second arms; and offsetting the first, second and third delays relative to said time origin of the switching fronts of three of said three first arms and said three second arms that are not blocked.

8. A double three-phase rotating electrical machine comprising an integrated inverter provided with a control device as claimed in claim 1.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 shows diagrammatically a double three-phase rotating electrical machine and an inverter provided with its control device in accordance with the invention.

(2) FIGS. 2a, 2b, 2c and 2d show timing diagrams of the pulse width modulation applied to the double three-phase machine, respectively resulting from first, second, third and fourth control strategies.

(3) FIG. 3 shows the operating points of the double three-phase machine shown in FIG. 1 for the selection of the control strategies shown in FIGS. 2a, 2b, 2c and 2d.

(4) FIG. 4 shows a selection tree for the control strategies shown in FIGS. 2a, 2b, 2c and 2d as a function of the operating points shown in FIG. 3.

DESCRIPTION OF PREFERRED EMBODIMENTS OF THE INVENTION

(5) As the diagrammatic representation of FIG. 1 clearly shows, in a preferred embodiment of the invention the stator 1 of the double three-phase rotating electrical machine includes a star-connected first three-phase system consisting of three first phase windings 2, 3, 4 and a second star-connected three-phase system consisting of three second phase windings 5, 6, 7 offset relative to one another by an offset angle of 30.

(6) Each of the first phase windings 2, 3, 4 has a first end connected to each of the first mid-points of first power arms 8 formed by switching elements 9 of a first power module 10 and a common other first end 11.

(7) In the same way, each of the second phase windings 5, 6, 7 has a second end connected to each of the second mid-points of second power arms 12 formed by other switching elements 13 of a second power module 14 and another common second end 15.

(8) The common first and second ends 11, 15 are the neutral points of the first and second three-phase systems and are isolated from each other.

(9) The first and second power arms 8, 12 of the first and second power modules 10, 14 are connected in parallel to a direct current source CC that includes a decoupling capacitor 16.

(10) The switching elements 9, 13 are controlled by a control device 17 so as to switch first phase currents R, S, T circulating in the first phase windings 2, 3, 4 and second phase currents U, V, W circulating in the second phase windings 5, 6, 7 in accordance with control strategies implemented in the invention, making it possible to have function as a six-phase inverter the first and second power modules 10, 14 functioning as first and second three-phase inverters INV1, INV2.

(11) In each of these first and second inverters, it is possible to block a power arm 8, 12 or not. It is also possible to offset the PWM relative to one another.

(12) The inventors have determined that these control strategies make it possible both to reduce the switching losses in the switching elements 9, 13, most often consisting of MOS transistors, at the same time as reducing an RMS current that the decoupling capacitor 16 at the input of the inverter 10, 14 must absorb.

(13) Four control strategies studied by the inventors by means of computer simulations based on work carried out on three-phase machines are described in detail hereinafter with reference to FIGS. 2a, 2b, 2c and 2d.

(14) Strategy I

(15) This first strategy consists in applying two vector PWM without blocking a power arm 8, 12.

(16) The timing diagrams of the first phase currents R, S, T generated by the first three-phase inverter INV1 and the second phase currents U, V, W generated by the second three-phase inverter INV2 in this first strategy are shown in FIG. 2a.

(17) A centered vector PWM (SVM) is applied to each of the two three-phase systems.

(18) The PWM have the same period. They are separated by a constant or variable common delay Td.

(19) Strategy II

(20) This second strategy consists in applying two vector PWM with blocking of one or both power arms 8, 12.

(21) The timing diagrams of the first phase currents R, S, T generated by the first three-phase inverter INV1 and the second phase currents U, V, W generated by the second three-phase inverter INV2 in this second strategy are shown in FIG. 2b.

(22) The vector PWM with an offset of the potentials of the neutral points 11, 15 make it possible to block one or two of the six power arms 8, 12 of the inverter 10, 14.

(23) The PWM have the same period. They are separated by a constant or variable common delay Td.

(24) Simulations have shown that the rms current in the decoupling capacitor 16 is greatly reduced when this common delay Td is 25% of the PWM period.

(25) Strategy III

(26) This third strategy consists in applying two generalized discontinuous PWM (GDPWM).

(27) The timing diagrams of the first phase currents R, S, T generated by the first three-phase inverter INV1 and the second phase currents U, V, W generated by the second three-phase inverter INV2 in this third strategy are shown in FIG. 2c.

(28) The two GDPWM are applied to each of the two three-phase systems, that is to say: blocking in each of the two three-phase inverters INV1, INV2 of the power arm 8, 12 having a maximum current and able to be blocked; two of the remaining PWM are offset.

(29) The PWM have the same period. They are separated by a constant or variable common delay Td.

(30) Simulations have shown that the current in the decoupling capacitor 16 is greatly reduced when this common delay Td is 30% of the PWM period.

(31) Strategy IV

(32) This strategy consists in applying two centered vector PWM (SVM) to offset PWM and blocked power arm 8, 12.

(33) The timing diagrams of the first phase currents R, S, T generated by the first three-phase inverter INV1 and the second phase currents U, V, W generated by the second three-phase inverter INV2 in this fourth strategy are shown in FIG. 2d.

(34) The PWM have the same period.

(35) Two of the six power arms 8, 12 are blocked. Three of the four PWM are then offset by first, second and third delays Td1, Td2, Td3 the values of which are included in the range [50%, +50%] of the PWM period.

(36) These offsets are more general than for centered vector PWM (for which the delays are zero) and GDPWM (for which the delays equal 50).

(37) In accordance with the object of the invention, the control strategies described above lead to a reduction of the losses in the switching elements 9 and a reduction of the rms current in the decoupling capacitor 16.

(38) The inventors having noted that the losses in this decoupling capacitor 16 depending on the phase difference between a phase current R, S, T; U, V, W and a phase voltage and the coefficient of modulation m (defined as the ratio of a peak phase voltage to half a supply voltage of the inverter), there existed an additional optimization approach through dynamically selecting a current strategy from the set of control strategies as a function of an operating point of the polyphase machine.

(39) The typology of the various operating points (motor torque r, rotation speed ) taken into account for the selection of the control strategies is indicated in FIG. 3:

(40) Area A

(41) This first area A corresponds to starting the double three-phase machine 1 operating at constant motor torque 0 from stopped (zero rotation speed ) up to a first rotation speed 1.

(42) In an application of the double three-phase machine 1 as a starter generator in a motor vehicle, this first area A typically corresponds to the MOTOR mode of a STOP/START function.

(43) The rotation speed and the phase voltage of the machine 1 are low.

(44) Area B

(45) This second area B corresponds to a mode of operation of the machine 1 in which the rotation speed is between the first rotation speed 1 and a second rotation speed 2 that marks the end 18 of operation at constant torque.

(46) In an application of the double three-phase machine 1 to assisting acceleration in a hybrid vehicle, this second area B typically corresponds to a BOOST function.

(47) The rotation speed and the phase voltage of the machine 1 remain low, but the power factor cos of the machine 1 may be low for a rotation speed close to the first rotation speed 1 or close to unity for a rotation speed close to the second rotation speed 2.

(48) Area C

(49) This third area C corresponds to a mode of operation of the machine 1 in which the rotation speed is between the second rotation speed 2 and a third rotation speed 3 that is representative of operation 19 at constant power.

(50) In an application of the double three-phase machine 1 to assisting acceleration in a hybrid vehicle, this third area C typically corresponds to a BOOST function.

(51) The rotation speed and the phase voltage of the machine 1 are high, and the power factor cos of the machine 1 is close to unity.

(52) Area D

(53) This fourth area D corresponds to a mode of operation of the machine 1 in which the rotation speed is greater than the third rotation speed 3.

(54) In an application of the double three-phase machine 1 to assisting acceleration in a hybrid vehicle, this fourth area D typically corresponds to a BOOST function.

(55) The rotation speed of the machine 1 is very high and the phase voltage is high, and the power factor cos of the machine 1 is close to unity.

(56) The rotation speed of the machine 1 being very high in this fourth area D, a current strategy of PWM type is replaced by full-wave control.

(57) This fourth area D therefore does not constitute a criterion for selection of a current strategy from the control strategies in accordance with the invention.

(58) FIG. 4 shows the tree for selection of control strategies as a function of the area A, B, C or D.

(59) On starting the machine 1, from stopped 20, the second strategy 21 is selected by a first test 22 for as long as the operating points of the machine 1 are in the first area A.

(60) In this first area A, the advantage of this second strategy 21 is to lower both the rms current in the decoupling capacitor 16 and the losses in the switching elements 9, 13. The switching losses are optimized.

(61) When a second test 23 shows that the operating points of the machine 1 are in the second area B, the second strategy 21 or the third strategy 24 is selected.

(62) The choice 25 between the second strategy 21 and the third strategy 24 depends on the power factor cos of the machine 1 relative to a predetermined coefficient preferably equal to 0.7.

(63) In this second area B, the second strategy 21 is selected if the power factor cos is less than or equal to 0.7 whereas the third strategy 24 is selected if the power factor cos is greater than 0.7.

(64) The reduction of the rms current in the decoupling capacitor 16 is not optimized in this second area B but is large for all the operating points. The switching losses are reduced.

(65) When a third test 26 shows that the operating points of the machine 1 are in the third area C, the third strategy 24 is selected.

(66) In this third area C, the reduction of the rms current in the decoupling capacitor 16 is close to the optimum. The switching losses are also optimized.

(67) When a fourth test 27 shows that the operating points of the machine 1 are in the fourth area D, none of the control strategies in accordance with the invention is selected. As already indicated, full-wave control 28 is therefore used.

(68) As goes without saying, the invention is not limited to only the preferred embodiments described above.

(69) An infinite number of control strategies is possible. The inventors have developed during computer simulations two further examples of vector PWM the parameters of which depend on the operating points of the machine:

Example 1

(70) Two vector PWM are applied using the fourth strategy described above (FIG. 2d): blocking in each of the two three-phase inverters INV1, INV2 of the power arm 8, 12 in which the absolute value of the current is maximum and that can be blocked; three of the remaining four PWM are offset in accordance with a first set Set1 and a second set Set2 of values of the first, second and third delays Td1, Td2, Td3 indicated in Table I below as a percentage of the PWM period; the first set Set1 is chosen if the phase difference is between 0 and /6 (modulo /3), otherwise the second set Set2 is chosen.

(71) TABLE-US-00001 TABLE I Delays Td1 Td2 Td3 Set1 40% 20% 40% Set2 40% 30% 30%

(72) In this first example, the reduction in the rms current in the decoupling capacitor 16 is the optimum for the operating points of the machine 1 at which the power factor is close to unity (cos =+/1) and/or the coefficient of modulation m is greater than 0.8. For the other operating points the reduction is small.

(73) In this first example, the switching losses for their part are optimized for all the operating points.

(74) In an application of the double three-phase machine 1 in a hybrid vehicle, this first example is advantageously used in the third area C, which typically corresponds to the BOOST function.

Example 2

(75) Two generalized discontinuous PWM (GDPWM) are applied in a similar manner to the third strategy described above (FIG. 2c) with a common delay Td between the first and second inverters INV1, INV2 depending on the phase difference and the coefficient of modulation m. A map of this common delay Td (as a percentage of the PWM period) is given in Table II below:

(76) TABLE-US-00002 TABLE II 0 30 36 45 60 75 90 1 20% 20% 25% 50% 50% 0 0 0.88 25% 10% 10% 0 0 0 0 0.7 40% 0 0 0 40% 0 0 0.5 40% 20% 20% 20% 25% 0 0 0.2 25% 25% 25% 25% 25% 0 0

(77) In this second example, the reduction of the rms current in the decoupling capacitor 16 is high for the operating points of the machine 1 at which the power factor is close to unity (cos =+/1) and/or the coefficient of modulation m is close to 1 (m=1).

(78) The control device 17 in accordance with the invention advantageously includes a Freescale MPC5643L microcontroller 29 including six independently programmable PWM registers that therefore facilitate implementation of the control strategies described above.

(79) The tests carried out by the inventors show that the control device in accordance with the invention, with the control strategies used, leads to a reduction in the rms current in the decoupling capacitor 16 as high as 63% compared to a basic PWM on a single three-phase machine.

(80) The preferred embodiments of the invention concern a double three-phase machine having an offset angle equal to 30. The same analysis could be performed for double three-phase rotating electrical machines having different offset angles .

(81) The control device in accordance with the invention employs microprogrammed logic 29 or alternatively hardwired logic or a programmed system or even an analog system.

(82) The invention therefore encompasses all possible variant embodiments to the extent that those variants remain within the scope defined by the following claims.