WO2015067593A1

WO2015067593A1 - Method and device for operating a permanently excited synchronous machine

Info

Publication number: WO2015067593A1
Application number: PCT/EP2014/073675
Authority: WO
Inventors: Peter Stauder; Tom Kaufmann; Sighard SCHRÄBLER
Original assignee: Continental Teves Ag & Co. Ohg
Priority date: 2013-11-06
Filing date: 2014-11-04
Publication date: 2015-05-14
Also published as: DE102013222539A1

Abstract

The invention relates to a method for operating a permanently excited synchronous machine, wherein a torque-generating current (Iq) and a magnetic field-weakening current (Id) are regulated in a rotor-fixed coordinate system. The aim of the invention is to allow an accurate and precise regulation even in the event of quick and dynamic processes. This is achieved in that in response to a specified desired torque, a synchronous machine operating point in which the magnetic field-weakening current (Id) has the lowest possible value is determined with a corresponding value of the torque-generating current (Iq).

Description

description

Method and device for operating a permanently excited synchronous machine

The invention relates to a method for operating a permanent-magnet synchronous machine in which a torque-forming current and a magnetic field weakening current are regulated in a rotor-fixed coordinate system. It also relates to a corresponding device.

A permanent magnet synchronous machine (PMSM) can be used to increase the torque achieved and thus also to increase the power achieved in the field weakening range. The need for this arises from the fact that with increasing speed in the phase windings an ever greater counter voltage is induced, so that the achievable with the predetermined supply voltage speed is limited. The description of the currents in the phase windings of the stator is usually carried out in a rotorfesten, d. H. With the rotor co-rotating coordinate system by means of a torque-generating current Iq and a magnetic field generating or magnetic field weakening current Id. Due to the field weakening, which corresponds to the condition Id <0, a component in

Phasor diagram of the synchronous machine generates the induced voltage or the electromotive force (EMF)

is opposite and thus reduces the terminal voltage. Usually, the proportion of the magnetic field weakening current Id is determined via a regulator which uses as input the current amplitude of the terminal voltage, which results as the square root of the sum of the squares of the voltages Ud and Uq, ie as (Ud2 + Uq2) l / 2, and the available DC voltage or voltage is used. A regulator operating in this manner is referred to as field weakening regulator and is known in the art. The disadvantage of such a control of the synchronous machine is that with very fast, dynamic

Speed changes only insufficient performance is achieved, especially in low-inertia systems.

From the prior art is also known to perform the control of the synchronous machine by means of maps, whereby a very dynamic control of the field weakening current Id is made possible. The disadvantage here is that the parameters of the motor and the inverter due to temperature influences (magnetic flux, resistance) and the state of the electrical system (voltage, allow power reference and feedback) vary and therefore either a variety of maps tables is required or corresponding deviations must be tolerated , by which the accuracy of such a control suffers. The invention is therefore based on the object to provide a method for operating a synchronous machine, which allows accurate and precise control even with fast and dynamic operations. Furthermore, a corresponding device should be specified.

With regard to the method, the abovementioned object is achieved according to the invention in that, for a given torque request, an operating point of the synchronous machine is determined with an associated value of the torque-forming current in which the current weakening the magnetic field has the lowest possible value. Advantageous embodiments of the invention are the subject of the dependent claims.

The invention is based on the consideration that a deposit of current quantities, in particular of Id and Iq, which are valid in a rotor-fixed coordinate system, in the form of characteristic curves or tables only insufficiently takes into account the current operating conditions. In order to cover as many operating conditions as possible, a large number of tables to be created in advance would be necessary, which is only rarely feasible in rare cases.

As has now been recognized, modern engine control units or control and regulating units (ECU) allow a determination of the operating point optimized for a torque request, adapted to the specific operational and dynamic driving situation. This determination can be made by the execution of a numerical method in which, in particular iteratively, values of Iq and Id are determined which are particularly suitable for realizing the desired torque.

Advantageously, the value of the magnetic field weakening current is determined iteratively. Preferably, the value of the torque-forming current is determined iteratively. This allows in each case a gradual improvement of the determined solution. By choosing the maximum number of iterations, higher accuracy can be set at the expense of speed. In a preferred embodiment, determined values for the torque-generating and the magnetic-field-weakening current are calculated as a resulting electrical power, with the aid of this power and the available battery voltage a required current is calculated, and wherein the value for the magnetic field weakening current is latched as a valid value when the required current is within predetermined limits. When calculating the electrical power, internal mechanical power and copper losses are preferably taken into account.

When calculating the internal mechanical power and an internal torque, preference is given to the reluctance of the machine, if present.

If a found value for the magnetic field weakening current would result in exceeding the available battery voltage, a gradient of the change in the necessary phase voltage is advantageously formed, in which case the gradient is greater than zero the value of the magnetic field weakening current in the next iteration diminished, otherwise increased. The step size for the iterative determination of the magnet ^¬ field-weakening current depends advantageously from the difference between a minimum possible and a maximum possible value for the magnetic-field-weakening current. The maximum number of iterations for the calculation of the magnetic field weakening current is preferably between 2 and 20, in particular at 10.

With regard to the device, the above-mentioned object is achieved according to the invention with at least one software and / or hardware implemented module for performing a method according to one of the above claims. The inventive apparatus is preferably used in a motor vehicle, where in particular in a brake ^¬ system, more preferably in a brake system of the 'brake-by-wire ". It can be used both in synchronous machines in linear actuators in an electro-mechanical Brake press a brake pad against a brake disc, as well as with linear actuators in integrated

Braking systems that move in a pressure supply device a plunger in a hydraulic pressure chamber, wherein active pressure can be built in at least one brake circuit.

Thus, the invention comprises a brake system for a motor vehicle with at least one electromechanical brake, which comprises a linear actuator, in particular comprising an electric motor and a downstream rotational-translation gear, preferably a ball screw, and a device described above. It also includes a braking system for a

Motor vehicle with a hydraulic Druckbereitstellungs- device with a linear actuator, in particular comprising an electric motor and a downstream Rotati ^¬ ons translation gear, preferably a ball screw, and a device described above. The advantages of the invention are in particular that a fast and at the same time flexibly adaptable to the current operating conditions control and regulation of the engine is made possible by a numerical or mathematical determination of the optimized operating point of the engine. By an iterative determination of the magnetic field weakening current, the optimized and, with the other parameters, the smallest possible compatible value can be determined. An embodiment of the invention will be described with reference to a

Drawing explained in more detail. Therein, the sole figure shows a structogram for a method for operating a permanent-magnet synchronous machine in a preferred embodiment. A combination of the torque-forming current Iq and the magnetic-field-weakening current Id is iteratively searched to a predetermined desired torque or torque command Msoll corresponding to a lowest possible magnetic field ^¬ debilitating value of Id. The method is preferably implemented as a computer program and runs on a motor ^¬ control unit or a control unit (ECU).

First, a method for reluctance torque motors in a preferred embodiment will be described with reference to the figure. Then, with reference to the figure, the changes that result for the method when performed for motors without reluctance torque will be discussed. The reference numbers, which characterize method steps or decisions and blocks, are entered in the respective fields in the structogram.

The method uses the following input parameters which characterize the electrical machine, and the concrete implementation of the driving Ver ^¬:

Psim: temperature-dependent main flow of the electric machine for phase as RMS value

Kt, inv, fak: reciprocal of the temperature-dependent torque constant for the setpoint current calculation

Ld inductance in D-axis

Lq inductance in Q-axis

P pole pair number

Ubat DC voltage or DC voltage of the battery

board network Ns: speed

Ibat, min: minimum battery current (eg for feeding back into the electrical system, for example -20 A)

Ibat, max: maximum battery current (at power consumption, eg 60 A)

Imax: Inverter current limit in Arms (RMS current)

Id, minimal, fak: factor for determining the minimum current Id for utilization of the reluctance torque

Ubat, fak: maximum degree of modulation (max Uac / Udc, eg 95%) nid, max: maximum number of iterations for Id

niq, max: maximum number of iterations for Iq

Rs: resistance for the engine phase, including the share of the ECU

Brake: Flag for increased copper losses in braking mode, which reduces the return power to the battery

These are inputs to the process, some of which can be adapted to the specific dynamic driving and / or controllers specific situ ^¬ ation. Thus, for example, Ubat, Ibat, min, Ibat, max and Imax can be adapted to the current vehicle electrical system conditions before carrying out the method. By setting the number of maximum iterations, the accuracy of determining Id and Iq can be set. The factor Id, minimum, fak is calculated at the beginning of the procedure as follows: first, a quantity D is formed according to (Ld-Lq) * Imax. If D <0, two auxiliary quantities H and G are formed according to

H = Psim / D

G = -asin (1/4 * (- (H2 + 8) 1/2-H))

This invoice must not necessarily be carried out as part of the United ^¬ proceedings or in the Reglerloop when the Inductances Ld and Lq are sufficiently stable. They change significantly and they are by means of a model in Ab ^¬ dependence of the currents Id and Iq (actual values from the last loop) calculated, this calculation should be carried out for 5 cycles.

If D> = 0, the value of G is set to zero, i. H. There is no Reluktanzmoment at negative values of Id. Even at low speeds, therefore, no negative current Id 10 should be impressed in order to obtain an increased torque.

This case arises for example with surface magnets (Ld = Lq) or with machines with "salient pole behavior" (Ld> Lq).

For Iq, max we get Imax * cos (G). The factor Id, 15 minimal, fak then results in

Id, minimal, fak = Imax * sin (G) / Iq, max2.

Furthermore, the torque command Msoll is important for determining the operating point.

20

IINN eeiinneemm eerrsstteenn BBlloocckk 22 wwiirrdd ooff ooff SSoollllssttrroommwweerrtt IIqq ffrroomm ddeemm DDrreehhmmoommeennttwwuunnsscchh bbeerreecchhnneett ggeemmääßß IIqq ,, ssoollll == MMssoollll ** KKTT ,, ,, iinnvv ffaakk .. IINN eeiinneerr EEnnttsscchheeiidduunngg 66 wwiirrdd ddaannnn üübbeerrpprrüüfftt ,, tthhee oobb ssoo eerrmmiitttteellttee WWeerrtt ddeeffeecctt IIqq ggrröößßeerr iisstt aallss ooff ooff nneeggaattiivvee WWeerrtt IImmaaxx .. IIsstt ddiieess 2255 tthhee FFaallll ,, wwiirrdd ooff ooff WWeerrtt IIqq ,, ssoollll iinn eeiinneemm BBlloocckk 88 aauuff ddeenn WWeerrtt ooff IImmaaxx ggeesseettzztt .. AAnnddeerrnnffaallllss wwiirrdd IIqq ,, ssoollll iinn eeiinneemm BBlloocckk 1122 aauuff ddeenn WWeerrtt ooff - llmmaxx ggeesseettzztt ((Iinn eeiinneerr E'nntsundschedinigduunngg 1100 wwiirrdd ddaabbeeii üübbeerrprprüüfftt oobb IIqq ,, ssoollll << == --IImmaaxx ggiilltt)).

3300 IInn eeiinneemm BBlloocckk 1166 wweerrddeenn nnuunn eeiinniiggee ffüürr ddiiee ffoollggeennddeenn IItteerraattiioonneenn nnoottwweennddiiggee GGrröossßeenn * Id, min = 0: during this iteration of Id, the lowest value of Id to date is recorded at the corresponding operating point of the motor;

Id, hold = 0: after completion of the iterations, this value is assigned the iteratively determined value for Id;

Iq, hold: after the iterations have been completed, this value is assigned the iteratively determined value for Iq;

Ub = Ubat * 0.40824829 * Ubat, fak: equals the maximum available voltage (in [volts]) multiplied by a utilization factor.

Here, Ub applies = Ubat / 2/3 * fmod, where fmod a Exploitation ^¬ or modification factor. Division by 3 is used to convert phase-to-phase voltage to phase, and division by 2 is used to convert to an RMS value.

The modulation factor takes into account that the available DC link voltage of z. B. 12 V by means of a

Pulse width modulation (PWM) with a duty cycle of z. B. 5% to 95% can be applied to a motor phase. This corresponds to a degree of modulation of 90%. The causes lie in the power electronics.

This means that the maximum voltage that can be applied to a motor phase measured in rms voltage corresponds to the value of Ub. If the voltage is to be greater than Ub, the value of Ubat can not be used or the process does not work anymore, since the peak value of the outer conductor voltage is then greater than Ubat. Only below Ub can sinusoidal voltages be generated. Other values to initialize are:

Ubat, inv = 1 / Ubat: inverse value of the battery voltage

Ubat

Ubat2 = Ub2: length of the maximum allowed voltage vector ws = ns * 0.10471976: speed in the unit [rad / s], with a conversion factor from rpm to rad / s nlq = niq, max: number of iteration steps to determine Iq

nid = nid, max: number of iteration steps to determine Id

Uemk = ws * Psim * p: corresponds to the induced voltage in Vrms; Vrms here corresponds to the effective value of the induced voltage in a motor phase with respect to the neutral point

Iq = Iq, shall * 0.5: The value of Iq is set to half of the maximum value of Iq

Iq, step = 2 * Iq: this corresponds to the jump distance for the iteration of Iq

Imax2 = Imax2: the current limit of the inverter in square; this corresponds to the length of the maximum possible current vector simvalid = 0: flag, whether the simulation resp. their result is valid

breakiq = 0: Flag if the iteration should be aborted via Iq

Ibat, min, limit = Ibat, min + 0.01

In a block 20, a first iteration loop is started via the current Iq, which is carried out as long as the conditions nlq> 0 and breakiq are equal to zero. First, the current iteration number nlq is counted down by one. For a given value of Iq, an efficiency optimal value Id, minimum for Id calculated according to - (Iq2) * Id, minimal, fak. A jump distance Id, step for the iteration over the value of Id is determined according to Id, step = (Id, max - Id, minimal). The initial value for Id is now placed in the middle between the values Id, minimum and Id, max according to Id = Id, minimum + Id, step * 0.5. A variable idvalid is set to zero, indicating whether a valid Id value was found. Furthermore, a flag breakid is set to zero, which indicates whether it should continue to iterate over the value of Id. The value of Iq, step is now reduced to its half, and nid is set to nid, max.

In a block 24, a loop now begins over the values of Id, this loop being performed as long as the conditions nid> 0 and breakid are equal to zero. The iteration counter nid is now counted down by one. The

Increment Id, step is reduced to half of its current value. In addition, a squared total current Iges in the rotor-fixed coordinate system (q, d) is calculated according to Iges = Id2 + Iq2.

In a decision 28 it is checked whether the square of the total current, Iges, is greater than Imax2. If this is the case, the current values of Iq and Id do not correspond to a valid solution. The method then branches to a block 32 in which the current value of Id is reduced by the value Id, step.

On the other hand, if Iges <= Imax2, then the method branches in a block 36 in which various voltages occurring in the synchronous machine are calculated:

Urd = Id * Rs: Voltage drop at the resistor in d-axis Urq = Iq * Rs: Voltage drop at the resistor in q-axis

Uld = ws * Lq * Iq * p: Voltage drop at the

q-inductance of the d-axis

Ulq = ws * Ld * Id * p: voltage drop at the

d-inductance of the q-axis

Uq = Uemk + Urq + Ulq: Sum of the voltage vectors in the q-axis

Ud = Urd - Uld: Sum of the voltage vectors in the d-axis

Uges2 = + Ud2 Uq2: (squared) length of the Ge ^¬ samtspannungszeigers

In a decision 40 it is now checked whether the squared length of the total voltage vector Uges2 is greater than the square of the battery voltage Ubat2. If this is the case, there is no valid value for Id since the battery voltage represents the maximum available voltage. In this case, the method branches to a block 44 in which a gradient degree is formed according to degrees = (Id * (Rs2 + (ws * Ld * p) 2) + ws * p * (Ld * Uemk + Iq * Rs * (Ld - Lq))) The gradient degree formed in this way indicates the change of the necessary phase voltage. If, for example, the gradient is positive, this means that an increase in the magnitude of the current Id (eg from 60 A to 65 A) produces a reduction of the necessary voltage.

If the gradient is negative, the voltage required would be greater by stronger field ^¬ attenuation. The sign of degrees now determines in which direction the value of Id should be changed for the next iteration. If grad> 0, the new value Id + Id, step is assigned in a block 42 Id, otherwise the new value Id-Id, step is assigned in a block Id.

If decision 40 has shown that Uges2 <= Ubat2, i. h., the available DC voltage Ubat is not exceeded in the combination of the values of Iq and Id, the method branches to a block 48. In this, the occurring powers are calculated:

Pmech = 3 * p * Iq * (psim + (Ld - Lq) * Id) * ws

Pcu = (Id2 + Iq2) * Rs * 3;

Pel = Pmech + Pcu

Here, Pmech refers to the internal mechanical power, Pcu the copper losses, and Pel the required electrical power, which corresponds to the current values of Iq and Id. A current Ibat, test corresponding to this electric power results in Ibat, test = Pel * Ubat, inv.

In a decision 52 it is now checked whether the current Ibat, test is within the permitted limits of Ibat, max and Ibat, min. If this is the case, d. H. if Ibat, min <= Ibat, test <= Ibat, max, the method branches to a block 56. The flags simvalid and idvalid are set to the value 1 there, and Id, min is assigned the current value of Id. This value corresponds to the currently best solution for the current Id.

Regardless of the outcome of the decision

52, the method is continued in a decision 58. There, the simultaneous existence of the conditions Brake is equal to 1, ws * Msoll <0, and Ibat, test <Ibat, min, limit checked. If all three conditions are present, the method branches to a block 62 in which the current value of Id is increased by the value Id, Step. In a subsequent decision 66 it is checked whether the value of Id is smaller than the value of -Imax. If this is the case, in a block 70 the value of Id is set to the value of -Imax. If the three above-mentioned conditions are not present, the method branches to a block 74, in which Id is reduced by the value of Id, Step.

Independently is checked in a decision 78 from the result of decision 52, whether the value of Id is larger than the value of Id, minimal, multiplied by ^¬ a decrement factor, which in the present case 0.999. If this is the case, the flag breakid is set to 1 in a block 82.

In a decision 86 it is checked whether the flag idvalid has the value 1. If this is not the case, no valid value for Id was found in the current iteration. The method in this case branches to a block 90 in which the value of Iq is reduced by the value of Iq, step. In a subsequent decision 94 it is checked whether the sign of the product of Iq and Iq, step is negative. If this is the case, breakiq is set to the value one in a block 98.

If decision 86 indicates that idvalid is 1, a valid Id value was found. In a block 88, the value of idvalid is now reset to zero, and Id, hold is assigned the value of Id, min, which corresponds to the best value so far. In addition, Iq, hold the current value of Iq is assigned to ^¬ . The internal torque of the electric machine Ms is now calculated in block 88 according to FIG Ms = 3 * p * Iq * (psim + (Ld - Lq) * Id, min), where the internal mechanical power Pmech is Pmech = Ms * ws. The copper losses Pcu arise too

Pcu = (Id, min2 + Iq2) * Rs * 3, and the electric power is now calculated to Pel = Pmech + Pcu. The to this electrical power and the available DC voltage Ubat

Corresponding current Ibat, test is calculated according to Ibat, test = Pel * Ubat, inv. In a decision 102 is now tested whether at least one of the following conditions is met:

abs (Ms)> abs (Msoll)

• Ibat, test> Ibat, max

• Ibat, test <Ibat, min

The function abs () denotes the formation of the

Absolute amount of the value in brackets. The About ^¬ test in decision 102 can for example be performed such that the above three conditions are linked with a logical OR and any resulting expression is evaluated to determine whether he - in the boolean sense - "true" or "false". If the evaluation of the expression "true" results, the value of Iq is reduced by the value Iq, step in a block 106. Otherwise, in a block 110, the value of Iq is increased by the value Iq, step Decision 114 checks whether abs (Iq)> abs (Iq, soll), ie whether the value of Iq is greater than its setpoint If this is the case, breakiq is set to the value one in a block 120. In a decision Finally, it is checked whether the sign of the product of Iq with Iq, step is less than 0. If this is the case, breakiq is set to the value one in a block 128. The process steps performed in a final block 134 are performed after the two iterations over Iq and Id. The iterations were either terminated because the maximum number of iterations was reached, or because the breakiq or breakid flags were set to one, which is a non-zero value.

The variable Id is assigned the value of Id, hold, the variable Iq the value of Iq, hold. As in block 36, the voltages are calculated according to

Urd = Id * Rs

Urq = Iq * Rs

Uld = ws * Lq * Iq *

Ulq = ws * Id * Id *

Uq = Uemk + Urq + Ulq

Ud = Urd - Uld

The currents Iq, Id can now be used in the motor drive to set the desired optimized operating point of the motor. The described method is advantageously carried out whenever a new request for a desired torque occurs.

A method in a second preferred embodiment for motors without reluctance torque is explained below with reference to the same figure. Essentially, the differences to the method described above are shown. The sequence of the procedure (in the sense of the associated structogram) for motors without reluctance torque is identical to the version for motors with reluctance torque. Differences arise in ^¬ We sentlichen in that now no longer up Reluktanzterme occur, which in some places leads to simplifications of the procedure.

In block 20, Id, minimum is not calculated to calculate Id, Step, and Id. Rather, Id, Step are assigned the value -Imax and Id the value Id, Step * 0.5.

The mechanical power Pmech in block 48 is now calculated according to

Pmech = 3 * p * Iq * psim * ws. The contribution proportional to (Ld-Lq) is missing here, since the inductances Ld and Lq are equal in the considered motor without reluctance torque.

Also in the calculation of the gradient degree in block 44, the term proportional to (Ld-Lq) disappears, so that degree is now calculated according to

grad = (Id * (Rs2 + (ws * Ld * p) 2) + ws * p * (Ld * Uemk)). Another change occurs in block 88 in the calculation of the internal torque; this is given in the reluctance-free case by Ms = 3 * p * Iq * psim.

, ₀

Reference sign list

2 block

6 decision

8 block

12 block

16 block

20 block

24 block

28 decision

32 block

26 block

40 decision

42 block

44 block

46 block

48 block

52 decision

56 block

58 decision

62 block

66 decision

70 block

74 block

78 decision

82 block

86 decision

88 block

90 block

94 decision

98 block

102 decision

106 block , _

110 block

114 decision

120 block

124 decision

128 block

134 block

Claims

claims

1. A method for operating a permanent magnet synchronous machine in which in a rotor-fixed coordinate system, a torque-generating current (I _q ) and a magnetic field weakening current (I _d ) are controlled, characterized gekennzeichne that for a given torque request, an operating point of the synchronous machine with a supplied ^¬ impaired value of the torque-forming current (I _q ) is determined in which the magnetic field weakening current (I _d ) has the lowest possible value.

2. The method of claim 1, wherein the value of the magnetic field weakening current (I _d ) is determined iteratively.

3. The method according to claim 1 or 2, wherein the value of

torque-forming current (I _q ) is determined iteratively.

4. Method according to one of claims 1 to 3, wherein calculated values for the torque-forming (I _q ) and the magnetic field weakening current (I _d ) a resulting electrical power (P _e i) is calculated, and with the aid of this power (P _e i) and the available Batte ^¬ riespannung (U _bat), a required current (I _bat t, test) is calculated, and the value for the magnetic field ^¬ weakening current (I _d) is latched as a valid value when the required current (Ibatt, test) within ^¬ within predetermined limits.

5. The method according to claim 4, wherein in the calculation of the electrical power (P _e i) an internal mechanical power (P _meC h) and copper losses (P _cu ) taken into account ^¬ the. Method according to one of claims 1 to 4, wherein in the calculation of the internal mechanical power (P _meCh ) and an internal torque (M _s ), the reluctance of the synchronous _{machine is} taken into account.

Method according to one of claims 1 to 6, wherein, if a determined value for the magnetic field weakening current (I _d ) would lead to exceeding the available battery voltage (U _b at), a gradient of the change of the necessary phase voltage is formed, and wherein in the case that the gradient is greater than zero, the value of the magnetic field weakening current (I _d ) is reduced in the next iteration, otherwise increased.

Method according to one of claims 2 to 7, wherein the step size for the iterative determination of the magnet ^¬ field weakening current (I _d ) depends on the difference between a minimum possible and a maximum possible value for the magnetic field weakening current (I _d ).

Method according to one of claims 2 to 8, wherein the maximum number of iterations for the calculation of the magnetic field weakening current between 2 and 20, in particular at 10, is located.

Apparatus for operating a permanent-magnet synchronous machine in which a torque-generating current (I _q ) and a magnetic field weakening current (I _d ) are controlled in a rotor-fixed coordinate system with at least one software and / or hardware implemented module for carrying out a method according to one the previous claims.