OPTIMIZED CURRENT SPECIFICATION FOR AN EXTERNALLY IGNITED SYNCHRONOUS MOTOR

Description

field of technology

[0001] The present invention relates to an operating method for a separately excited synchronous motor comprising an excitation winding and a motor winding, wherein a setpoint determiner receives an instantaneous speed of the synchronous motor and a target torque to be supplied by the synchronous motor, wherein the setpoint determiner determines a setpoint for an excitation current to be supplied to the excitation winding and a motor current to be supplied to the motor winding, wherein the setpoint determiner specifies the determined setpoints as setpoints to a current control device for a converter device, so that the current control device controls the converter device in such a way that the converter device supplies the excitation current to the excitation winding and the motor current to the motor winding.

[0002] The present invention further relates to a computer program for a setpoint determiner, wherein the computer program comprises machine code that can be executed by the setpoint determiner, wherein the execution of the machine code by the setpoint determiner causes the setpoint determiner to perform such an operating procedure.

[0003] The present invention further relates to a setpoint determiner for determining setpoints for an excitation current and a motor current of a separately excited synchronous motor, wherein the setpoint determiner is programmed with such a computer program so that it performs such an operating procedure in operation.

[0004] The present invention further assumes a drive, wherein the drive comprises a separately excited synchronous motor with an excitation winding and a motor winding, wherein the drive comprises a converter unit which is connected to the excitation winding for supplying an excitation current and to the motor winding for supplying a motor current, wherein the drive comprises a current control unit which controls the converter unit, wherein the drive comprises such a setpoint determiner, wherein the setpoint determiner has inputs for receiving an instantaneous speed of the synchronous motor and a setpoint torque to be supplied by the synchronous motor, wherein the setpoint determiner is connected to the current control unit for specifying a value for a field-generating component of the motor current and a value for a torque-generating component of the motor current and a value for the excitation current. State of the art

[0005] In recent years, the use of externally excited synchronous motors (FESM) has increased significantly. This is due to their high efficiency across the entire operating range and their high starting torque. A further advantage, particularly compared to permanent magnet synchronous motors (PESM), is that no materials with limited availability are required, especially rare earth elements. Furthermore, the rotor's magnetic flux, adjustable via the excitation current, provides an additional degree of freedom, allowing for energy-efficient and flexible achievement of the desired torque.

[0006] The torque control of a FESM (Fiber-Electrically Adjustable Motor) is typically performed in field-oriented coordinates. This requires calculating target currents based on the specified torque, the motor parameters, the inverter data, and the current speed. Key requirements for these target currents are that the required torque is achieved as effectively as possible while minimizing losses. The calculation must be completed within a single control cycle and therefore be capable of real-time operation. Physical limitations, such as the maximum available voltage of the inverter and the maximum current in the motor, must be observed and therefore considered in the calculation.

[0007] In the prior art, the current components of the motor current, i.e., the field-generating and torque-generating components, as well as the excitation current, are determined partially independently of one another. The required torque is fed to a current filter, which uses filtering to determine a setpoint for the torque-generating current component of the motor current (usually referred to as the q-current) and provides this setpoint to the current control device. Furthermore, the maximum value of the motor voltage and the current motor voltage are fed to a field weakening controller. The field weakening controller determines the setpoint for the excitation current and provides this setpoint to the current control device. Additionally, the field weakening controller determines a preliminary setpoint for the field-generating component of the motor current (usually referred to as the d-current).Based on the current speed of the synchronous motor, a correction value for the setpoint of the field-forming component of the motor current is determined using a characteristic curve. The final setpoint for the field-forming component of the motor current is determined by adding this correction value to the preliminary setpoint. The final setpoint for the field-forming component of the motor current is then fed back to the current control unit.

[0008] Other approaches to FESM address this problem, but they only ever solve partial problems. Furthermore, these solutions are not real-time capable. Other state-of-the-art methods only consider the base speed range (where operation without field weakening is possible) and use characteristic curves for the field weakening controller in the field weakening range, i.e., for high speeds. These methods therefore require characteristic curves that must be recorded beforehand. Moreover, the motor current and the excitation current are considered independently, which produces suboptimal solutions.

[0009] It is also known in the art to use offline-optimized look-up tables for the currents to find an optimal operating point depending on the current speed and the requested torque. However, this solution requires a very large amount of measurement data and considerable preparation effort. The look-up tables for the motor current and the excitation current created in this way are inflexible during runtime and must be completely adjusted if the machine parameters change. Furthermore, the look-up tables are very memory-intensive if they are to be used with high accuracy and depending on the motor parameters.

[0010] From DE 10 2014 223 014 A1, an operating method for a separately excited synchronous motor is known, which has an excitation winding and a motor winding. In this operating method, a setpoint determiner receives the instantaneous speed of the synchronous motor and the target torque to be supplied by the synchronous motor. The setpoint determiner is aware of maximum values ​​for the excitation current supplied to the excitation winding, the motor voltage driving the motor current, the power of the synchronous motor, and also the resistance values ​​of the excitation winding and the motor winding. The setpoint determiner sets up an optimization problem for a current vector and solves it in real time. The current vector has a component each for a field-generating component of the motor current, a torque-generating component of the motor current, and the excitation current.The setpoint generator determines the current vector as part of the optimization problem solution in such a way as to minimize losses occurring in the excitation winding and the motor winding. The setpoint generator determines the current vector such that the actual torque of the synchronous motor, resulting from the excitation current and the motor current, corresponds to the target torque. As a supplementary condition, the setpoint generator considers that the magnitude of the excitation current does not exceed the maximum value for the excitation current, and the magnitude of the motor voltage does not exceed the maximum value for the motor voltage. The setpoint generator specifies the components of the determined current vector as setpoints to a current control device for a converter device, so that the current control device controls the converter device in such a way that the converter device supplies the excitation current to the excitation winding and the motor current to the motor winding.

[0011] US 2008 / 001570 A1 discloses an operating method for a separately excited synchronous motor having an excitation winding and a motor winding, wherein the motor current and motor voltage must always remain below their maximum values ​​and at low speeds the maximum torque is achieved at maximum excitation current. Summary of the invention

[0012] The state of the art does not lead to optimal operation of the separately excited synchronous motor in all operating conditions. In particular, field weakening and the corresponding determination of a correction value for the setpoint of the field-generating component of the motor current generally only occur when the speed of the separately excited synchronous motor is above a limiting speed.

[0013] For a PESM (Power-Electrical Synchronous Motor), methods exist for determining the optimal operating mode. However, the complexity and solvability of the problem are considerably greater for a FESM (Fiber-Electrical Synchronous Motor) than for a PESM. This also applies to the decision logic for finding the optimal solution. The object of the present invention is to provide methods for operating a separately excited synchronous motor in an optimized manner. In particular, losses of the synchronous motor are to be minimized. The solution should be capable of real-time operation.

[0014] The problem is solved by an operating method with the features of claim 1. Advantageous embodiments of the operating method are the subject of dependent claims 2 to 6.

[0015] According to the invention, an operating method for a separately excited synchronous motor is created, in which A setpoint generator receives an instantaneous speed of the synchronous motor and a target torque to be applied by the synchronous motor; the setpoint generator is aware of maximum values ​​for an excitation current supplied to the excitation winding, a motor current supplied to the motor winding, a motor voltage driving the motor current, and a power rating of the synchronous motor; the setpoint generator is also aware of resistance values ​​of the excitation winding and the motor winding; the setpoint generator checks whether a requested power of the synchronous motor, given by the product of the instantaneous speed and the required target torque, exceeds the maximum power rating; if the requested power does not exceed the maximum power rating, the setpoint generator sets up and solves a first optimization problem for a current vector in real time, and / or if the requested power exceeds the maximum power rating,A second optimization problem for the current vector is set up and solved in real time. The current vector has a component for a field-forming component of the motor current, a torque-forming component of the motor current, and the excitation current. The setpoint determiner determines the current vector within the framework of solving the first optimization problem in such a way that losses occurring in the excitation winding and the motor winding are minimized. Within the framework of the first optimization problem, the setpoint determiner considers as a supplementary condition that an actual torque of the synchronous motor resulting from the excitation current and the motor current corresponds to the setpoint torque. Within the framework of the first optimization problem, the setpoint determiner further considers first general boundary conditions according to which the magnitude of the motor current reaches a maximum of the maximum value for the motor current.-- the magnitude of the excitation current reaches a maximum of the maximum value for the excitation current and -- the magnitude of the motor voltage reaches a maximum of the maximum value for the motor voltage, the setpoint determiner determines the current vector within the framework of the solution of the second optimization problem such that the resulting actual torque of the synchronous motor is maximized, the setpoint determiner considers as a supplementary condition within the framework of the second optimization problem that the magnitude of the motor voltage is equal to the maximum value for the motor voltage, the setpoint determiner further considers second general boundary conditions within the framework of the second optimization problem, according to which -- the magnitude of the motor current reaches a maximum of the maximum value for the motor current and -- the magnitude of the excitation current reaches a maximum of the maximum value for the excitation current,and the setpoint determiner specifies the components of the current vector of a current control device, determined by solving the first or the second optimization problem, as setpoints for a converter device, so that the current control device controls the converter device in such a way that the converter device supplies the excitation current to the excitation winding and the motor current to the motor winding.

[0016] This results in a unified solution for all relevant currents (field-generating component, torque-generating component, excitation current). This allows the synchronous motor to operate with minimal losses in every operating condition where the target torque can be achieved. Conversely, in every operating condition where the target torque cannot be achieved, the synchronous motor can be operated in such a way that its torque is approximated as closely as possible to the target torque.

[0017] The solution according to the invention is therefore based on determining the motor current and the field current by solving a suitable optimization problem.

[0018] The first optimization problem—at least in real time and with currently available, mass-market hardware—is not solvable in its general form using a closed-form solution. Therefore, to efficiently solve the first optimization problem, it is preferably intended that the setpoint determiner, within the framework of solving the first optimization problem, The current vector is first provisionally determined without considering the first general boundary conditions, then it is checked whether the provisionally determined current vector satisfies the first general boundary conditions, the provisionally determined current vector is used as the current vector if the provisionally determined current vector satisfies the first general boundary conditions, and otherwise the current vector is determined taking into account at least one of the following first specific boundary conditions, according to which -- the magnitude of the motor current is equal to the maximum value for the motor current, -- the magnitude of the excitation current is equal to the maximum value for the excitation current, and -- the magnitude of the motor voltage is equal to the maximum value for the motor voltage, and depending on which of the first general boundary conditions are not satisfied, it is determined which of the first specific boundary conditions it takes into account.

[0019] The problem is thus divided into subproblems that can be solved analytically or at least approximately solvable numerically. In cases that cannot be solved analytically, efficient numerical optimizations are used, which lead to the optimal operating point in just a few iterations. Based on a decision tree, it is determined which of the subproblems will actually be solved and whose solution will then be used. This approach significantly reduces the complexity of the solution determination, thus ensuring real-time capability.

[0020] Preferably, it is provided that the setpoint determiner, as part of the examination to determine whether the preliminary determined current vector meets the first general boundary conditions, First, it checks whether the magnitude of the motor current reaches the maximum value for the motor current and whether the magnitude of the excitation current reaches the maximum value for the excitation current, and only then does it check whether the magnitude of the motor voltage reaches the maximum value for the motor voltage.

[0021] This approach quickly leads to the correct solution, i.e., the current vector, using a simple, easy-to-work-through decision tree.

[0022] To efficiently solve the second optimization problem, it is preferably provided that the setpoint determiner, within the framework of solving the second optimization problem, - first provisionally determines the current vector without considering the second general boundary conditions, - then checks whether the provisionally determined current vector satisfies the second general boundary conditions, - uses the provisionally determined current vector as the current vector if the provisionally determined current vector satisfies the second general boundary conditions, and otherwise determines the current vector taking into account at least one of the following second specific boundary conditions, according to which -- the magnitude of the motor current is equal to the maximum value for the motor current and -- the magnitude of the excitation current is equal to the maximum value for the excitation current, and - depending on which of the second general boundary conditions are not satisfied, determines which of the second specific boundary conditions it takes into account.

[0023] This significantly reduces the complexity of the investigation, thus ensuring real-time capability.

[0024] Preferably, it is provided that the setpoint determiner, as part of the examination of whether the preliminary determined current vector meets the second general boundary conditions, First, it checks whether the magnitude of the motor current reaches the maximum value for the motor current, and only then does it check whether the magnitude of the excitation current reaches the maximum value for the excitation current.

[0025] This approach quickly leads to the correct solution, i.e., the current vector, using a simple, easy-to-work-through decision tree.

[0026] Preferably, the setpoint determiner considers only copper losses as losses. While limiting the analysis to copper losses is a simplification, it is justified because copper losses constitute the majority of total losses (at least 70%, often 80% or more), and furthermore, synchronous motors are typically operated in their base speed range, where copper losses are particularly dominant over iron losses.

[0027] The problem is further solved by a computer program with the features of claim 7. According to the invention, the execution of the computer program by the setpoint determiner causes the setpoint determiner to execute an operating method according to the invention.

[0028] The problem is further solved by a setpoint determiner with the features of claim 8. According to the invention, the setpoint determiner is programmed with a computer program according to the invention, so that the setpoint determiner executes an operating procedure according to the invention during operation.

[0029] The problem is further solved by a drive with the features of claim 9. According to the invention, in a drive of the type mentioned at the outset, the setpoint determiner is designed as a setpoint determiner according to the invention. Brief description of the drawings

[0030] The properties, features, and advantages of this invention described above, as well as the manner in which they are achieved, will become clearer and more readily understandable in connection with the following description of the exemplary embodiments, which are explained in more detail in conjunction with the drawings. These drawings show, in schematic representation: FIG 1 a drive, FIG 2 a flowchart, FIG 3 a first optimization problem, FIG 4 a flowchart, FIG 5 a second optimization problem and FIG 6 a flowchart. Description of the embodiments

[0031] According to FIG 1 The drive system comprises a separately excited synchronous motor 1. The synchronous motor 1 has a stator 2 with a motor winding 3 arranged therein. A motor current I is supplied to the motor winding 3 via a converter 4. The motor winding 3 is generally designed as a three-phase winding. The motor current I therefore includes the phases a, b, c of a three-phase system. The synchronous motor 1 also has a rotor 5 with a co-rotating excitation winding 6. An excitation current Ie is supplied to the excitation winding 6 via another converter 7. The excitation current Ie is generally a direct current. The converter 4 and the converter 7 together form a converter unit 8. The converter unit 8 is controlled by a current control unit 9.

[0032] The drive also includes a setpoint determiner 10. The setpoint determiner 10 provides the current control unit 9 with setpoints consisting of a field-generating component Id and a torque-generating component Iq of the motor current I and the excitation current Ie. The field-generating component Id and the torque-generating component Iq together constitute the setpoint for the motor current I. The two components Id and Iq are spatially rotated by 90° relative to each other. Therefore, the motor current I is a vector quantity, as far as the internal handling of the motor current I in the setpoint determiner 10 and the specification of the setpoints to the current control unit 9 are concerned, where the following relationship applies in the (co-rotating) dq system: Id 2 + Iq 2 = I 2 applies.

[0033] Based on the specified setpoint values ​​Id, Iq, and Ie, the current control unit 9 is able to control the inverter unit 8 such that the inverter 7 supplies the excitation current Ie to the excitation winding 6 and the inverter 4 supplies the motor current I to the motor winding 3. The conversion from the rotating dq system to the abc system is familiar to experts and does not require further explanation.

[0034] The setpoint determiner 10 is the actual core subject matter of the present invention. The setpoint determiner 10 is programmed with a computer program 11. The computer program 11 comprises machine code 12, which can be executed by the setpoint determiner 10. Due to the execution of the machine code 12 by the setpoint determiner 10, the setpoint determiner 10 performs an operating procedure, which is described below, first in conjunction with FIG 2 , which will be explained in more detail with further reference to the other FIGs.

[0035] According to FIG 2 The setpoint determiner 10 is informed of the values ​​Imax, Iemax, and Umax in a single step S1. These values ​​Imax, Iemax, and Umax can, for example, be determined by the computer program 11 or set once during the commissioning of the drive. If necessary, they can also be determined dynamically depending on a state of the drive, for example, depending on the operating temperature of the synchronous motor 1.

[0036] The value Imax is the maximum value for the magnitude of the motor current I. The value Iemax is the maximum value for the magnitude of the excitation current Ie. The value Umax is the maximum value for the magnitude of the motor voltage U driving the motor current I. The maximum value Umax for the motor voltage U can be determined, for example, by the supply voltage of the inverter 4, such as an intermediate circuit voltage applied to the input side of the inverter 4.

[0037] The motor voltage U in the co-rotating dq system – analogous to the motor current I – also exhibits a d- and a q-component, hereinafter referred to as the field-forming and torque-forming voltage components Ud and Uq. Analogous to the motor current I, the following relationship applies: Ud 2 + Uq 2 = U 2

[0038] The setpoint determiner 10 is further informed in step S2 of the resistance R of the motor winding 3 and the resistance Re of the excitation winding 6. The values ​​R and Re can also be determined by the computer program 11 or set once during the commissioning of the drive.

[0039] Based on the now given values, the components Ud, Uq of the motor voltage U and an excitation voltage Ue, and the components Id, Iq of the motor current I and the excitation current Ie, can be related to each other. This is because, in steady-state operation, the following approximation holds true: Ud = R ⋅ Id − ω ⋅ Lq ⋅ Iq Uq = R ⋅ Iq + ω ⋅ Ld ⋅ Id + Lm ⋅ Ie Ue = Re ⋅ Ie

[0040] In equations 3 to 5, ω is the electrical angular frequency. Ld and Lq are the self-inductances of motor winding 3 along the d- and q-axes, respectively. Lm is the mutual inductance of the excitation winding 6.

[0041] Equation 5 also indirectly shows that the maximum excitation voltage Ue does not need to be explicitly considered, as it is linearly related to the excitation current Ie. The maximum value Iemax can therefore be determined such that it corresponds to the earlier limiting factor (excitation current Ie or excitation voltage Ue).

[0042] Equations 3 to 5 do not account for the effects of changes in the motor current I and the excitation current Ie over time. This is permissible because these changes and their resulting effects are small. Furthermore, equations 3 to 5 assume that the resistances R and Re are constant over time. However, any temperature dependence or dependence on other drive conditions can easily be included.

[0043] The self-inductances Ld and Lq of motor winding 3 and also the mutual inductance Lm of excitation winding 6 depend on the components Id and Iq of the motor current I and the excitation current Ie. Where necessary, the specific values ​​for the self-inductances Ld and Lq of motor winding 3 and the mutual inductance Lm of excitation winding 6 can be stored in lookup tables within the setpoint determiner 10.

[0044] The setpoint determiner 10 is further informed of a maximum value Pmax for a power P of the synchronous motor 1. It is possible that the maximum value Pmax for the power P of the synchronous motor 1 is explicitly specified to the setpoint determiner 10. Alternatively, it is possible that the setpoint determiner 10 itself determines the maximum value Pmax in a step S3 according to the relationship P max = U max ⋅ Im ax − R ⋅ Im ax 2 determined.

[0045] In step S4, the setpoint determiner 10 receives an instantaneous (mechanical) rotational speed n of the synchronous motor 1 and a target torque M* to be applied by the synchronous motor 1. The method of specifying the target torque M* can be of any nature. It is not, as such, the subject of the present invention. Likewise, the method of specifying the instantaneous rotational speed n is not, as such, the subject of the present invention. For example, the drive can have a detection block 13 that determines the rotational speed n based on the actual operating state of the synchronous motor 1. Alternatively, instead of the rotational speed n, the electrical angular frequency ω can be determined and the rotational speed n calculated from it. Corresponding procedures for determining the electrical angular frequency ω and also for calculating the rotational speed n from the electrical angular frequency ω are generally known to those skilled in the art.Furthermore, it is assumed that the target torque M* has a positive value. A negative value for the target torque M* would only affect the sign; the magnitudes for the field-generating component Id, the torque-generating component Iq, and the excitation current Ie would remain unchanged.

[0046] In step S5, the setpoint determiner calculates the currently requested power P of the synchronous motor 1. The currently requested power P of the synchronous motor 1 is given by the product of the current speed n and the required target torque M*.

[0047] In step S6, the setpoint determiner 10 checks whether the currently requested power P exceeds the maximum value Pmax. If the requested power P does not exceed the maximum value Pmax, the setpoint determiner 10 proceeds to step S7. In step S7, the setpoint determiner 10 sets up a first optimization problem O1 for a current vector i and solves it in real time. If, however, the requested power P exceeds the maximum value Pmax, the setpoint determiner 10 proceeds to step S8. In step S8, the setpoint determiner 10 sets up a second optimization problem O2 for the current vector i and solves it in real time.

[0048] The current vector i comprises three components: the field-generating component Id, the torque-generating component Iq of the motor current I, and the excitation current Ie. The solutions to the first and second optimization problems O1 and O2 are therefore the values ​​that the setpoint determiner 10 specifies as setpoints for the current control device 9. This specification occurs in step S9.

[0049] It is possible to use the procedure of FIG 2 It should be added that if the setpoint determiner 10 solves the second optimization problem O2, it issues a message to a higher-level system (not shown). This allows the higher-level system to be informed that the requested setpoint torque M* cannot be provided.

[0050] As part of solving the first optimization problem O1, the setpoint determiner 10 determines according to FIG 3 The current vector i is adjusted such that the losses V that occur are minimized. Within the scope of the present invention, only the copper losses VK occurring in the excitation winding 6 and the motor winding 3 are considered. The other losses – in particular the iron losses – can be neglected, since in practice the copper losses VK are considerably larger than the other losses.

[0051] The copper losses VK can be approximated to a good degree in the form of the relationship VK = 3 2 ⋅ R ⋅ Id 2 + Iq 2 + Re ⋅ Ie 2 The first optimization problem O1 is applied. Within the framework of the first optimization problem O1, the setpoint determiner 10 considers as a supplementary condition the fundamental prerequisite of the first optimization problem O1 that an actual torque M of the synchronous motor 1 corresponds to the setpoint torque M*. The actual torque M is determined by the excitation current Ie and the components Iq, Id of the motor current I. This results in a good approximation of M = 3 2 Z ⋅ Lm ⋅ Ie + Ld − Lq ⋅ Id ⋅ Iq Z is the number of pole pairs of the synchronous machine 1.

[0052] Furthermore, the target value determiner 10 considers first general boundary conditions within the framework of the first optimization problem O1. According to these boundary conditions, it is required that the magnitude of the motor current I reaches a maximum value Imax for the motor current I, the magnitude of the excitation current Ie reaches a maximum value Iemax for the excitation current Ie, and the magnitude of the motor voltage U reaches a maximum value Umax for the motor voltage U.

[0053] Mathematically, the first optimization problem O1 can therefore be formulated as follows: i = arg min i V taking into account Id 2 + Iq 2 ≤ Im ax 2 Ie ≤ Ie max Ud 2 + Uq 2 ≤ U max 2 M = M *

[0054] Solving the first optimization problem O1 is not a trivial task, since the actual moment M depends in a complex way on the components of the current vector i, and furthermore, the nonlinear constraints according to inequalities 10 to 12 must be taken into account. The exact procedure for solving the first optimization problem O1 is described below in conjunction with FIG 4 explained in more detail.

[0055] FIG 4 shows the specific procedure by which the setpoint determiner 10 determines the current vector i in the context of solving the first optimization problem O1.

[0056] According to FIG 4 The setpoint determiner 10 first determines the current vector i=i1 in step S11. The setpoint determiner 10 determines the current vector i1 without considering the first general boundary conditions. Thus, it does solve the first optimization problem O1 according to equation 9, taking equation 13 into account. However, it solves the optimization problem O1 without simultaneously considering inequalities 10 to 12.

[0057] In practice, step S11 can be solved analytically using a Lagrange approach. This solution is known from the literature. For example, the article "Optimal current control of externally excited synchronous machines in automotive traction drive applications" by O. Haala, B. Wagner, M. Hofmann, and M. Marz, published in the International Journal of Electrical and Computer Engineering, Volume 7 (2013), pages 1133 to 1139, can be cited. The textbook "Electric Drives: Control of Drive Systems" by D. Schroder, Springer-Verlag, 2009, can also be mentioned. The solution is independent of the rotational speed n or the electrical angular frequency ω. It is known in technical circles as Maximum Torque Per Current (MTPC).

[0058] The solution determined in step S11 is only provisional. Before the setpoint determiner 10 adopts the current vector i=i1 as the final current vector i, it first checks in step S12 whether the magnitude of the motor current I reaches the maximum value Imax for the motor current I. If this condition is met, the setpoint determiner 10 then checks in step S13 whether the magnitude of the excitation current Ie reaches the maximum value Iemax for the excitation current Ie.

[0059] If the check in step S12 is already negative, meaning that the magnitude of the motor current I for the current vector i1 is greater than the maximum value Imax for the motor current I, the setpoint determiner 10 determines the current vector i=i2 in a step S14. The setpoint determiner 10 determines the current vector i2 as before by solving equation 9, taking equation 13 into account. However, it solves equation 9, additionally also considering the first special boundary condition. Id 2 + Iq 2 = Im ax 2 This boundary condition is taken into account. Due to this constraint, the term to be minimized within Equation 9 can be significantly simplified. This is because the term R(Id 2< +Iq 2< ) is a constant according to Equation 14, with the value RImax 2< and can therefore be neglected during the minimization. Consequently, the resistance Re can also be neglected, as it represents only a constant factor. Instead of minimizing the (total) copper losses VK, it is therefore sufficient to minimize the term Ie 2<. Strictly speaking, it is even sufficient to minimize the magnitude of the excitation current Ie. Furthermore, according to Equation 14, the magnitude of the moment-generating component Iq can be determined from the value of the field-generating component Id, so that one less variable needs to be varied. In practice, step S14 can also be solved analytically using a Lagrange approach.

[0060] If the current vector i passes the test of step S12 but fails the test of step S13, the setpoint determiner 10 determines the current vector i=i3 in a step S15. The setpoint determiner 10 determines the current vector i3 as before by solving equation 9, taking equation 13 into account. However, it solves equation 9 additionally, taking the first special boundary condition into account. Ie = Ie max taken into account.

[0061] Step S15 has as its solution the MTPC trajectory of a permanent magnet synchronous machine. This solution is known from the literature. For example, reference can be made to the paper "Analytical solutions for the optimal reference currents for MTPC / MTPA, MTPV and MTPF control of anisotropic synchronous machines" by H. Eldeeb, C.M. Hackl, J. Kullick, and L. Horlbeck, published in Proc. 2017 IEEE International Electric Machines and Drives Conference (IEMDC), 2017, pages 1 to 6. Reference can also be made to the paper "Optimal setpoint computation for constrained torque control of PMSMs" by T. Englert and K. Graichen, published in Proc. 2018 European Control Conference (ECC), 2018, pages 2671 to 2677.

[0062] Analogous to the procedure for determining the current vector i=i2, the term to be minimized in Equation 9 for determining the current vector i=i3 can also be significantly simplified. This is because the term ReIe 2< is a constant according to Equation 15, with the value ReIemax 2<, and can therefore be neglected during the minimization. Consequently, the factor 3R / 2 can also be neglected, as it is simply a constant. Instead of minimizing the (total) copper losses VK, it is therefore sufficient to minimize the term Id 2< +Iq 2<. Furthermore, due to Equation 15, one less variable needs to be varied, since the magnitude of the excitation current Ie is fixed. The excitation current Ie can therefore only have the values ​​+Iemax or -Iemax.

[0063] The current vectors i=i2 and i=i3 from steps S14 and S15 are also only provisional. Before the setpoint determiner 10 adopts the current vector i=i2 or the current vector i=i3 as the final current vector i, it checks in steps S16 and S17, respectively, whether the remaining current condition is satisfied. This means that inequality 11 is satisfied in the case of current vector i=i2, and inequality 10 in the case of current vector i=i3. If, in the first case, current vector i=i2 does not satisfy inequality 11, or in the second case, current vector i=i3 does not satisfy inequality 10, the setpoint determiner 10 determines the current vector i=i4 in step S18. The current vector i4 is determined by the setpoint determiner 10 as before by solving equation 9, taking equation 13 into account. However, it solves equation 9 while simultaneously considering the conditions according to equation 14 and equation 15.

[0064] The solution to step S18 is (almost) trivial, since only the field-generating component Id needs to be varied in the result. The moment-generating component Iq is thereby fixed – except for its sign. Likewise, the excitation current Ie is also fixed except for its sign.

[0065] Alternative approaches to the implementation of steps S11 to S18 are also possible. In particular, it is possible to perform the checks of steps S12 and S13 in reverse order, thus primarily determining the current vector i=i3 rather than the current vector i=i2. An even better approach combines the checks of steps S12 and S13. If, within this configuration, the current vector i=i1 fails either of the two checks, the setpoint determiner 10 determines the two current vectors i=i2 and i=i3 and checks whether each of them passes the other check. The setpoint determiner 10 thus checks whether the current vector i=i2 satisfies the condition of inequality 11 and whether the current vector i=i3 satisfies the condition of inequality 10.If both current vector i=i2 and current vector i=i3 pass their respective tests, the setpoint determiner 10 selects as current vector i the one of the two current vectors i=i2 and i=i3 that exhibits the lower copper losses VK. If only one of the two current vectors i=i2 and i=i3 passes its test, while the other fails, the setpoint determiner 10 selects as current vector i the one of the two current vectors i=i2 and i=i3 that passes its test. If both current vectors i=i2 and i=i3 fail their respective tests, the setpoint determiner 10 continues with the determination of current vector i=i4.

[0066] The current vector i now determined satisfies the current limitations of inequalities 10 and 11, regardless of whether it is current vector i1, i2, i3, or i4. However, it is not yet guaranteed that the voltage limitation according to inequality 12 is also satisfied. Therefore, in step S19, the setpoint determiner 10 checks whether the determined current vector i fulfills the condition according to inequality 12. In other words, the setpoint determiner 10 checks whether the magnitude of the motor voltage U reaches a maximum of the maximum value Umax for the motor voltage U.

[0067] If the current vector i passes the test in step S19, the setpoint determiner 10 uses the current vector i determined by processing steps S11 to S18. Otherwise, the setpoint determiner 10 determines the current vector i=i5 in step S20. The setpoint determiner 10 determines the current vector i5 as before by solving equation 9, taking equation 13 into account. However, it solves equation 9 while simultaneously fulfilling the condition Ud 2 + Uq 2 = U max 2 taken into account.

[0068] To determine the current vector i=i5, it may be necessary to calculate it numerically. Various methods are available to those skilled in the art for this purpose.

[0069] For example, it is possible to solve the current vector i=i5 using gradient descent methods, line search methods, or multidimensional Newton-Raphson methods. However, these methods are prone to being too time-consuming. Due to the structure of the optimization problem for the current vector i=i5, analytical partial solutions can be determined, thus allowing the derivation of highly efficient numerical methods.

[0070] An example of such a method is fixed-point iteration for the excitation current Ie. In this case, the necessary first-order optimality conditions of the optimization problem are determined using the Lagrange approach. La i , λ 1 , λ 2 = i T ⋅ R ′ ⋅ i + λ 1 ⋅ i T ⋅ M ′ ⋅ i − M * + λ 2 ⋅ i T ⋅ U ′ ⋅ i − U max 2 can be determined. La is the Lagrangian. λ1 and λ2 are Lagrange multipliers. R', M', and U' are 3x3 matrices which can be determined from the resistances R, Re, the generated torque M, the voltage components Ud and Uq, and the excitation voltage Ue in conjunction with the boundary conditions to be observed (equations 13 and 16). The gradient ∂ La i , λ 1 , λ 2 ∂ = 2 R ′ + λ 1 ⋅ M ′ + λ 2 ⋅ U ′ ⋅ i The minimum condition must be met. R ′ + λ 1 ⋅ M ′ + λ 2 ⋅ U ′ ⋅ i = 0 This condition can be used to calculate the excitation current Ie and the Lagrange multipliers λ1 and λ2 as a function of the components Id and Iq of the motor current I. Specifically, the function exists Ie = f Ie Id Iq .

[0071] The solution to this function does not depend on the Lagrange multipliers λ1 and λ2.

[0072] Similarly, this can be applied to the boundary conditions for the torque M (equation 13) and the motor voltage U (equation 16). The field-generating component Id and the torque-generating component Iq of the motor current I can be expressed as functions of the excitation current Ie. Id = f Id Ie and Iq = f Iq Ie will be calculated.

[0073] Based on the functions f Ie , f Id and f Iq, an iterative solution method in the form of a fixed-point iteration can be used. Ie k + 1 = f Ie f Id Ie k , f Iq Ie k The formula is formulated where k is the respective iteration step. The fixed point of the iteration formula is then the optimal current vector i=i5, and the Lagrange multipliers λ1 and λ2 can be calculated analytically.

[0074] The iteration is terminated at a suitable point to guarantee real-time capability. The termination criterion could be, for example, that a maximum number of iterations have been performed or that the result is only changing marginally. By using a suitable initial approach, the actual solution is obtained in very few iterations. The initial approach could, for example, be the excitation current Ie last output by the setpoint determiner 10 to the current control device 9. However, other approaches are also possible.

[0075] Alternatively, a one-dimensional method can be used to determine a zero of a function. These methods are very efficient. In this case, it is assumed that equation 19 must have a nontrivial solution. Therefore, the matrix must R ′ + λ 1 ⋅ M ′ + λ 2 ⋅ U ′ The matrix must have a kernel, and its determinant must be zero. This allows expression 24 to be solved for λ1 or λ2. For example, λ1 can be expressed as a function of λ2. This yields the solutions for the optimal current vector i=i5, because the current vector i=i5 must be the kernel of the matrix for the gradient equation to be satisfied. Therefore, the following must hold: ker n R ′ + λ 1 ⋅ M ′ + λ 2 ⋅ U ′ = IL ⋅ i λ 2 .

[0076] Here, i(λ2) is a current vector of length 1, which thus determines the "direction" of the current vector i=i5, i.e., the ratio of the three components of the current vector i=i5. The length IL is determined via the conditions of equations 13 and 16. This is because the following must hold: IL 2 ⋅ i λ 2 T ⋅ R ′ ⋅ i λ 2 = M * and IL 2 ⋅ i λ 2 T ⋅ U ′ ⋅ i λ 2 = U max 2

[0077] Solving equations 26 and 27 for IL 2< and equating them yields the following equation: U max 2 ⋅ i λ 2 T ⋅ M ′ ⋅ i λ 2 − M * ⋅ i λ 2 T ⋅ U ′ ⋅ i λ 2 = 0 .

[0078] In equation 28, only λ2 is unknown. Therefore, equation 28 can be solved for λ2 using an ansatz. However, no analytical solution in closed form is known for equation 28. To determine the zero, i.e., the specific value of λ2, generally known methods for finding zeros can be used, such as Newton's method or the interval bisection method.

[0079] After numerically solving equation 28, the components Id, Iq of the motor current I and λ1 can be calculated analytically. Again, the optimum from the previous calculation step can be used for an initial solution.

[0080] Of course, other solution methods can also be used.

[0081] As part of the described procedures, it must also be ensured that the solution found lies within the permissible current range. To verify this, equations 13 (in conjunction with equation 8), 14, and 16 can be solved analytically. This yields a calculated minimum and a calculated maximum value for the excitation current Ie, which must be adhered to. If the calculated minimum value is less than 0, it is set to 0. If the calculated maximum value is greater than the maximum excitation current Imax, the maximum excitation current Imax is used instead of the calculated maximum value. If the excitation current Ie of the current vector i=i5 lies outside the operating range determined in this way, the calculated minimum value is the optimal solution.This ensures that the power restrictions are observed and that a valid solution can always be found which meets all restrictions and simultaneously provides the requested moment M*.

[0082] As part of solving the second optimization problem O2, the setpoint determiner 10 determines the current vector i according to FIG 5 such that the actual torque M of the synchronous motor 1 is maximized, meaning that the actual torque M is brought as close as possible to the target torque M*. The losses V, and in particular the copper losses VK, are not considered in the solution of the second optimization problem O2.

[0083] Within the framework of the second optimization problem O2, the setpoint determiner 10 considers as a supplementary condition that equation 16 is satisfied. This corresponds in substance to the aim of controlling the inverter 4 as much as possible so that the actual torque M of the synchronous motor 1 is maximized.

[0084] Furthermore, the target value determiner 10 takes into account second general boundary conditions within the framework of the second optimization problem O2. According to these boundary conditions, it is required that the magnitude of the motor current I reaches a maximum value Imax for the motor current I (inequality 10) and the magnitude of the excitation current Ie reaches a maximum value Iemax for the excitation current Ie (inequality 11).

[0085] Mathematically, the second optimization problem O2 can therefore be formulated as follows: i = arg min i − M where the setpoint determiner 10 also takes into account equation 16 and inequalities 10 and 11. The exact procedure for solving the second optimization problem O2 is described in conjunction with FIG 6 explained in more detail.

[0086] FIG 6 shows the specific procedure by which the setpoint determiner 10 determines the current vector i in the context of solving the second optimization problem O2.

[0087] According to FIG 6 The setpoint determiner 10 first determines the current vector i in step S31 as the current vector i6. The setpoint determiner 10 determines the current vector i6 without considering the second general boundary conditions. Thus, it solves the second optimization problem O2 according to equation 29, taking equation 16 into additional consideration. However, the setpoint determiner 10 does not consider inequalities 10 and 11 in step S31.

[0088] In practice, step S31 can be solved analytically using a Lagrange approach, for example. This solution can be performed analogously to determining the current vectors i2 and i3.

[0089] The solution determined in step S31 is only provisional. Before the setpoint determiner 10 adopts the current vector i6 as the final current vector i, the setpoint determiner 10 first checks in step S32 whether the magnitude of the motor current I reaches the maximum value Imax for the motor current I.

[0090] If the current vector i=i6 fails the test in step S32, the setpoint determiner 10 determines the current vector i as current vector i7 in step S33. In step S34, the setpoint determiner 10 again solves equation 29. However, it additionally considers not only equation 16, but also equation 14.

[0091] Then, in step S34, the setpoint determiner 10 checks whether the current vector i now determined - regardless of whether it is the current vector i6 or i7 - fulfills the condition that the excitation current Ie satisfies inequality 11, i.e., the magnitude of the excitation current Ie reaches a maximum of the maximum value Iemax for the excitation current Ie.

[0092] If the current vector i passes the test in step S34, the setpoint determiner 10 uses the current vector i on which the test was based, i.e., the current vector i=i6 or the current vector i=i7. If the current vector i fails the test in step S34, the setpoint determiner 10 determines the current vector i8 in step S35. In step S35, the setpoint determiner 10 again solves equation 29. However, it also considers not only equation 16 but also equation 15.

[0093] Analogous to the procedures for the current vectors i=i2 and i=i3, one less variable needs to be varied when determining the current vectors i=i7 and i=i8, because the magnitude of the moment-forming component Iq can be determined from the value of the field-forming component Id, or the magnitude of the excitation current Ie is fixed.

[0094] In step S35, another optimization problem must be solved. This optimization problem maximizes the torque M with the maximum value Iemax of the excitation current Ie while adhering to the voltage limit according to Equation 16. This optimization problem is known for permanent magnet synchronous machines and has an analytical solution there. The solution trajectory is referred to in the literature as Maximum Torque Per Voltage (MTPV) and is known, for example, from the previously mentioned documents "Analytical solutions for the optimal reference currents for MTPC / MTPA, MTPV and MTPF control of anisotropic synchronous machines" by H. Eldeeb et al. and "Optimal setpoint computation for constrained torque control of PMSMs" by T. Englert and K. Graichen. It can be adopted directly.

[0095] In practice, step S35 can be solved analytically using a Lagrange approach, for example.

[0096] In step S36, the setpoint determiner 10 checks whether the magnitude of the motor current I reaches the maximum value Imax for the current vector i=i8. If the current vector i passes the test in step S36, the setpoint determiner 10 uses the current vector i=i8 determined in step S35. If the current vector i fails the test in step S36, the setpoint determiner 10 determines the current vector i=i9 in step S37. In determining the current vector i9, the setpoint determiner 10 again solves equation 29. However, it also considers not only equation 16 but also equation 14.

[0097] This optimization problem is known for permanent magnet synchronous machines and has an analytical solution. The solution is known, for example, from the previously mentioned publications "Analytical solutions for the optimal reference currents for MTPC / MTPA, MTPV and MTPF control of anisotropic synchronous machines" by H. Eldeeb et al. and "Optimal setpoint computation for constrained torque control of PMSMs" by T. Englert and K. Graichen.

[0098] Even in the design of FIG 6 Modifications to the procedure are possible. These modifications are analogous to modifications to the design of FIG 4 In particular, it is possible to perform the tests of steps S32 and S34 in reverse order, thus primarily determining the current vector i=i8 rather than the current vector i=i7. An even better approach combines the tests of steps S32 and S34. If, within this configuration, the current vector i=i6 fails either of the two tests, the setpoint determiner 10 determines the two current vectors i=i7 and i=i8 and checks whether each of them passes the other test. The setpoint determiner 10 thus checks whether the current vector i=i7 satisfies the condition of inequality 11 and whether the current vector i=i8 satisfies the condition of inequality 10. If both the current vector i=i7 and the current vector i=i8 pass their respective tests, the setpoint determiner 10 selects as current vector i the one of the two current vectors i=i7 and i=i8 that provides the higher actual moment M.If only one of the two current vectors i=i7 and i=i8 passes its test, while the other fails, the setpoint determiner 10 selects the one that passes its test as current vector i. If both current vectors i=i7 and i=i8 fail their respective tests, the setpoint determiner 10 continues with the determination of current vector i=i9.

[0099] The current vectors i1 to i4 of FIG 4 as well as i6 to i9 from FIG 6can be calculated analytically and therefore very efficiently. While only a numerical solution is possible for the current vector i5, this solution can also be calculated very quickly, allowing the current vector i5 to be determined in real time. By dividing the two optimization problems O1 and O2 into several subproblems, which are solved depending on the situation (adherence to current and voltage limits, adherence to only one current limit, adherence to only the other current limit, etc.), real-time capability is achieved, particularly for determining the current vector i. The current vector i includes the field-generating component Id, the torque-generating component Iq of the motor current I, the excitation current Ie, and thus all relevant currents.

[0100] The present invention has many advantages. For example, the method according to the invention can be used with any FESM (Field-Electrical Motor Generator). No previously expensively recorded or pre-calculated look-up tables for the motor current I and / or the excitation current Ie are required. Furthermore, no critical simplifications are made. Only the motor parameters need to be known. This allows the method according to the invention to be quickly adapted for a new synchronous motor 1. Compared to methods that consider only a partial operating range of the FESM (in particular the base speed range) and then use a field weakening controller to comply with the operating limits, the method according to the invention offers the advantage that all operating ranges are considered in an energy-optimized manner. By calculating the truly optimal operating point, more torque is obtained at the same electrical power.The same torque is achieved with lower electrical power. By taking all current and voltage limitations into account, improved torque characteristics at high speeds are also obtained. The operating method according to the invention can be implemented and applied to all separately excited synchronous motors 1. Retrofitting existing drives is also possible. Compared to prior art solutions, the potential savings in electrical energy are expected to be at least 3% and can reach up to 10%.

[0101] Although the invention has been illustrated and described in detail by the preferred embodiment, the invention is not limited by the disclosed examples and other variants can be derived from them by a person skilled in the art - within the scope of the claims - without leaving the scope of protection of the invention. Reference symbol list

[0102] 1 Synchronous motor 2 Stator 3 Motor winding 4, 7 Inverter 5 Rotor 6 Excitation winding 8 Inverter unit 9 Current control unit 10 Setpoint determiner 11 Computer program 12 Machine code 13 Determination block a, b, c Phases i, i1 to i9 Current vectors I Motor current Id Field-forming component Ie Excitation current Iq Torque-forming component Imax Maximum value (motor current) Iemax Maximum value (excitation current) M Torque M*Target torque n Speed ​​O1, O2 Optimization problems P Power Pmax Maximum value (power) R Resistance (motor winding) Re Resistance (excitation winding) S1 to S37 Steps U Motor voltage Ue Excitation voltage Ud, Uq Components of the motor voltage Umax Maximum value (motor voltage) ωelectric angular frequency

Claims

1. Operating method for an externally excited synchronous motor (1) comprising an excitation winding (6) and a motor winding (3), - wherein a setpoint value determining unit (10) receives an instantaneous rotational speed (n) of the synchronous motor (1) and a setpoint torque (M*) to be applied by the synchronous motor (1), - wherein maximum values (Imax, Iemax, Umax) for an excitation current (Ie) supplied to the excitation winding (6), a motor current (I) supplied to the motor winding (3), a motor voltage (U) driving the motor current (I), and a power (P) of the synchronous motor (1) are known to the setpoint value determining unit (10), - wherein resistance values (Re, R) of the excitation winding (6) and the motor winding (3) are also known to the setpoint value determining unit (10), characterized in that - the setpoint value determining unit (10) checks whether a requested power (P) of the synchronous motor (1) given by the product of the instantaneous rotational speed (n) and the setpoint torque (M*) to be applied exceeds the maximum value (Pmax) for the power (P), - the setpoint value determining unit (10), in the case that the requested power (P) does not exceed the maximum value (Pmax) for the power (P), sets a first optimization problem (O1) for a current vector (i) and solves it in real time and / or in the case that the requested power (P) does exceed the maximum value (Pmax) for the power (P), sets a second optimization problem (02) for the current vector (i) and solves it in real time, - each current vector (i) has a component for a field-forming component (Id) of the motor current (I), a torque-forming component (Iq) of the motor current (I), and the excitation current (Ie), - the setpoint value determining unit (10) determines the current vector (i) in the context of the solution of the first optimization problem (O1) in such a way that losses (V) occurring in the excitation winding (6) and the motor winding (3) are minimized, - the setpoint value determining unit (10) takes the fact that an actual torque (M) of the synchronous motor (1) resulting from the excitation current (Ie) and the motor current (I) corresponds to the setpoint torque (M*) into account as a supplementary condition in the context of the first optimization problem (O1), - the setpoint value determining unit (10) also takes first general boundary conditions into account in the context of the first optimization problem (O1), in accordance with which -- the magnitude of the motor current (I) at most reaches the maximum value (Imax) for the motor current (I), -- the magnitude of the excitation current (Ie) at most reaches the maximum value (Iemax) for the excitation current (Ie), and -- the magnitude of the motor voltage (U) at most reaches the maximum value (Umax) for the motor voltage (U), - the setpoint value determining unit (10) determines the current vector (i) in the context of the solution of the second optimization problem (02) in such a way that the resulting actual torque (M) of the synchronous motor (1) is maximized, - the setpoint value determining unit (10) takes the fact that the magnitude of the motor voltage (U) is equal to the maximum value (Umax) for the motor voltage (U) into account as a supplementary condition in the context of the second optimization problem (02), - the setpoint value determining unit (10) also takes second general boundary conditions into account in the context of the second optimization problem (02), in accordance with which -- the magnitude of the motor current (I) at most reaches the maximum value (Imax) for the motor current (I), and -- the magnitude of the excitation current (Ie) at most reaches the maximum value (Iemax) for the excitation current (Ie), and - the setpoint value determining unit (10) specifies the components of the current vector (i) determined by solving the first or the second optimization problem (O1, 02) as setpoint values to a current control device (9) for a converter device (8), so that the current control device (9) actuates the converter device (8) in such a way that the converter device (8) supplies the excitation current (Ie) to the excitation winding (6) and the motor current (I) to the motor winding (3).

2. Operating method according to Claim 1, characterized in that, in the context of the solution of the first optimization problem (O1), the setpoint value determining unit (10) - initially provisionally determines the current vector (i) without taking into account the first general boundary conditions, - then checks whether the provisionally determined current vector (i) satisfies the first general boundary conditions, - uses the provisionally determined current vector (i) as the current vector (i) if the provisionally determined current vector (i) satisfies the first general boundary conditions, and otherwise determines the current vector (i) taking into account at least one of the following first specific boundary conditions, according to which -- the magnitude of the motor current (I) is equal to the maximum value (Imax) for the motor current (I), -- the magnitude of the excitation current (Ie) is equal to the maximum value (Iemax) for the excitation current (Ie), and -- the magnitude of the motor voltage (U) is equal to the maximum value (Umax) for the motor voltage (U), and, - depending on which of the first general boundary conditions are not satisfied, determines which of the first specific boundary conditions it takes into account.

3. Operating method according to Claim 2, characterized in that, in the context of checking whether the provisionally determined current vector (i) satisfies the first general boundary conditions, the setpoint value determining unit (10) - first checks whether the magnitude of the motor current (I) at most reaches the maximum value (Imax) for the motor current (I) and whether the magnitude of the excitation current (Ie) at most reaches the maximum value (Iemax) for the excitation current (Ie), and - only then checks whether the magnitude of the motor voltage (U) at most reaches the maximum value (Umax) for the motor voltage (U).

4. Operating method according to Claim 1, 2 or 3, characterized in that, in the context of the solution of the second optimization problem (02), the setpoint value determining unit (10) - initially provisionally determines the current vector (i) without taking into account the second general boundary conditions, - then checks whether the provisionally determined current vector (i) satisfies the second general boundary conditions, - uses the provisionally determined current vector (i) as the current vector (i) if the provisionally determined current vector (i) satisfies the second general boundary conditions, and otherwise determines the current vector (i) taking into account at least one of the following second specific boundary conditions, according to which -- the magnitude of the motor current (I) is equal to the maximum value (Imax) for the motor current (I), and -- the magnitude of the excitation current (Ie) is equal to the maximum value (Iemax) for the excitation current (Ie), and - depending on which of the second general boundary conditions are not satisfied, determines which of the second specific boundary conditions it takes into account.

5. Operating method according to Claim 4, characterized in that, in the context of checking whether the provisionally determined current vector (i) satisfies the second general boundary conditions, the setpoint value determining unit (10) - first checks whether the magnitude of the motor current (I) at most reaches the maximum value (Imax) for the motor current (I), and - only then checks whether the magnitude of the excitation current (Ie) at most reaches the maximum value (Iemax) for the excitation current (Ie).

6. Operating method according to one of the preceding claims, characterized in that the setpoint value determining unit (10) only takes the copper losses (VK) into account as losses (V).

7. Computer program for a setpoint value determining unit (10), wherein the computer program comprises machine code (12) which can be executed by the setpoint value determining unit (10), wherein the execution of the machine code (12) by the setpoint value determining unit (10) has the effect that the setpoint value determining unit (10) carries out an operating method according to any one of the preceding claims.

8. Setpoint value determining unit (10) for determining setpoint values for an excitation current (Ie) and a motor current (I) of an externally excited synchronous motor (1), wherein the setpoint value determining unit (10) is programmed with a computer program (11) according to Claim 7, so that it carries out an operating method according to any one of Claims 1 to 6 during operation.

9. Drive, - wherein the drive has an externally excited synchronous motor (1) with an excitation winding (6) and a motor winding (3), - wherein the drive has a converter device (8) which is connected to the excitation winding (6) for supplying an excitation current (Ie) and to the motor winding (3) for supplying a motor current (I), - wherein the drive has a current control device (9) which actuates the converter device (8), - wherein the drive has a setpoint value determining unit (10) according to Claim 8, - wherein the setpoint value determining unit (10) has inputs for receiving an instantaneous rotational speed (n) of the synchronous motor (1) and a setpoint torque (M*) to be applied by the synchronous motor (1), - wherein the setpoint value determining unit (10) is connected to the current control device (9) for the specification of a value for a field-forming component (Id) of the motor current (I) and a value for a torque-forming component (Iq) of the motor current (I) and a value for the excitation current (Ie).