Method for operating a three-phase motor

Abstract

The invention relates to a method for operating a three-phase motor (4) on a multi-phase electrical network (5) by means of a motor control unit (1) with which one or more network phases (a, b, c) of the network (5) can be switched by switching on power electronic actuators (3). In addition to a network-related switching criterion, a rotor flux-related switching criterion is taken into account.

Description

[0001] The invention relates to a method for operating a three-phase motor. The invention also relates to a motor control unit. Likewise, the invention relates to a computer program product with which the method can be carried out.

[0002] European patent application EP 3916998 A1 (Siemens AG) dated December 1, 2021, and the conference paper "Novel predictive start-up algorithm for soft starter driven induction motors," hereinafter referred to as "Nannen et al.", presented by Hauke ​​Nannen, Heiko Zatocil, and Gerd Griepentrog, published in IECON 2020, 46th Annual Conference of the IEEE Industrial Electronics Society (IES), October 18-21, 2020, Singapore, describe a model-predictive control method for asynchronous machines with soft starters, which can reduce motor and thyristor losses. Based on measured and / or calculated state variables such as mains voltage, mains frequency, rotor speed, and rotor flux, a real-time simulation predicts current and torque profiles for all possible thyristor ignition combinations. These profiles are then evaluated based on decision criteria to identify usable ignition options.One disadvantage of this method is that predicting motor behavior in real time requires high computing power; this leads to significantly higher costs for the processor or microcontroller.

[0003] The present invention was made in light of the prior art described above, the object of which is to define a method which, on the one hand, offers similar operating behavior to the model predictive approach, but on the other hand does not require such high computing power.

[0004] The problem is solved by a method according to claim 1. This method serves to operate a three-phase motor on a multi-phase electrical network. Operation is achieved using a three-phase motor control unit, with which one or more network phases can be switched by switching power electronic actuators (e.g., MOSFET, IGBT, JFET, etc.). These power electronic actuators must not only be actively switched on to transition from a high-impedance, current-blocking state to a low-impedance, current-conducting state (the so-called "switching on" of a power electronic actuator), but also actively switched off again to return from the low-impedance, current-conducting state to the high-impedance, current-blocking state (the so-called "switching off" of a power electronic actuator).This fundamentally distinguishes these power electronic actuators from the thyristors previously used in motor starters, which, although they must be actively triggered to enter a conducting state, automatically become current-blocking again at the subsequent zero-crossing of the current. Preferably, the power electronic actuators are switched off precisely at the zero-crossing of the current.

[0005] By using electronic actuators that can be switched off, the system achieves a significant advantage: the ability to interrupt a short circuit. With an engine control unit (ECU) that uses one or more thyristors, if a short circuit occurs while the thyristor is conducting, one can only watch as the current rises so dramatically that the thyristor is destroyed. In contrast, with an ECU equipped with a switchable electronic actuator, the current can be selectively cut off to prevent damage to the ECU.

[0006] The operating procedure takes into account at least one grid-related switching criterion and at least one rotor flux-related switching criterion. The operating procedure also takes into account at least one grid-related switching criterion; however, it does not take into account a rotor flux-related switching criterion. Here, "grid-related switching criterion" means that the switching-off time is a function of the current flowing from the electrical grid into the three-phase motor.

[0007] Either 2-phase or 3-phase switching is used, whereby in a space vector representation a 2-phase switching leads to a current vector that is fixed in its orientation and pulsating in its amplitude, and a 3-phase switching leads to a rotating current vector.

[0008] The invention is based on the idea that a decision to activate one or more power electronic actuators is made with reference to a grid-related activation criterion, e.g., the position of the voltage vector, and a rotor flux-related activation criterion, e.g., the position of the rotor flux vector. In this respect, the control method according to the invention differs fundamentally from conventional control methods in which the rotor flux generated in the three-phase machine has no influence on the decision of whether to activate one or more power electronic actuators.

[0009] The presented method requires no detailed motor parameters or predictions, but only a few angular operations and comparisons, and is therefore many times easier to implement in an industrial product than the known model predictive control method. Initial estimates showed that the method according to the invention requires only about 10% of the computing power of the model predictive control method. Furthermore, controlled operation is even possible with the new method.

[0010] According to the invention, both the grid-related switching criterion and the rotor flux-related switching criterion must be met for the power electronic actuators to be switched on. Each switching option is then checked to see if it meets both the grid-related switching criterion and the rotor flux-related switching criterion. Only if a switching option meets both the grid-related switching criterion and the rotor flux-related switching criterion can the switching option actually be used.

[0011] According to the invention, the grid-related switching criterion is fulfilled if, in a space vector representation, the current vector of a switching possibility of two or more grid phases lies within a switching angle range defined by a lower switching angle range limit and an upper switching angle range limit with respect to the grid voltage and / or the grid current. The lower switching angle range limit determines the switching point at which the current flow reaches its maximum length and the largest current amplitude. Analogous to the firing angle α in classical, thyristor-based control methods, the lower switching angle range limit can be specified, for example, via a characteristic curve or controlled taking the current amplitude into account. Consideration of the current rotor flux amplitude to limit the maximum possible torque is also possible.Since the lower limit of the switching angle range can be used to control the current and torque, its function is similar to that of the ignition angle α in conventional control methods. The lower limit of the switching angle range can assume a value from the following range: [-90°; upper limit of the switching angle range], measured with respect to the grid voltage vector. The upper limit of the switching angle range indirectly defines the shortest current flow that can be triggered by switching. The upper limit of the switching angle range can be chosen to ensure that, even in a grid where zero-crossing determination is difficult, the switching always behaves in such a way that the current flows in the desired direction and that unwanted switching with excessively high current amplitudes does not occur.

[0012] According to the invention, the rotor flux-related switching criterion is fulfilled if, in a space vector representation, the current vector of a switching possibility of two or more network phases lies within a flux and torque angle range whose range limits are defined with respect to the rotor flux. By selecting one of the range limits of the flux and torque angle range, areas can be separated in which the switching operations, on average, generate a positive and a negative torque. Depending on the selection of this range limit, for example, a slightly negative torque is permitted in favor of a current that strongly generates rotor flux. The other range limit of the flux and torque angle range can be used to define how much positive torque and how much rotor flux linkage is generated in the rotor: it is possible that a build-up of positive torque leads to a reduction in rotor flux and vice versa.The choice of this other range limit value must therefore be carefully examined. This parameter can be used for "flow guidance" or "flow control" during startup. The mean absolute rotor flux can thus be determined using this limit value. |Ψ→2S| can be influenced. If the motor is to be accelerated, the difference between the range limits of the flux and torque angle ranges can assume a value from the following definition range: [90°, 180°], measured with respect to the rotor flux vector.

[0013] According to a preferred embodiment of the invention, the grid-related disconnection criterion consists in the fact that the disconnection of a power electronic actuator capable of switching a grid phase occurs at the current zero crossing of that grid phase. Preferably, the current flowing from the electrical grid into the three-phase motor is measured in the respective grid phases, and an electronic actuator arranged in a grid phase is switched off at the current zero crossing, i.e., when the current in that grid phase is zero.

[0014] According to a preferred embodiment of the invention, the presence of an activation criterion is determined by means of a space vector representation of rotor flux. |Ψ→2S|, Mains voltage U→1S and stator current I→1S The system has been tested. Using a space vector representation, the relationships between the network quantities voltage and current, the rotor flux quantities, the switching angle range, and the flux and torque angle range can be clearly illustrated. The main difference between the present invention and conventional control methods lies in the fact that the rotor flux space vector is taken into account when selecting a switching point. The calculation of a network space vector, e.g., the voltage space vector, and the rotor flux space vector, as well as the switching angle and flux and torque angle ranges, can be performed cyclically. Each time a switching opportunity arises within the switching angle and flux and torque angle ranges, the power electronic actuators can be controlled. This results in switching pulses that consider both the network voltage and the rotor flux.In this way, a positive torque, a limited stator current and a sufficient rotor flux for subsequent switching operations can be obtained.

[0015] According to a preferred embodiment of the invention, the flux and torque angle range is shifted by a rotation angle to compensate for a time delay between a switching decision and the switching-on. This reduces the effect of estimations of positive torque and rotor flux development being based on a different assumption than what actually occurs during the switching-on due to the rotating rotor. The rotation by the rotation angle shown in the space vector representation is intended to compensate for the deviation caused by the rotating rotor as effectively as possible.

[0016] According to a preferred embodiment of the invention, the three-phase motor is an asynchronous motor, a synchronous motor or a PM line start motor (PM = permanent magnet).

[0017] According to a preferred embodiment of the invention, the testing of grid-related and rotor flux-related switching criteria is preceded by a time period of flux build-up in which rotor flux is generated by switching on the power electronic actuators.

[0018] According to a preferred embodiment of the invention, only those switching options are identified that result in negative torques and thus actively brake the three-phase motor. Therefore, the method can also be used for active braking.

[0019] The problem is also solved by an engine control unit according to the invention, which is suitable for carrying out the steps of the method according to the invention. The engine control unit comprises a switching signal unit for generating switching signals and the power electronic actuators. The switching signal unit can be controlled to generate switching signals for one or more of the power electronic actuators. The engine control unit also comprises means suitable for carrying out the steps of the described method. These means can, for example, be a control unit which includes a processing unit and a storage unit. A computer program can be loaded into the storage unit and permanently stored there, from where it can be loaded into the processing unit for execution.

[0020] The outlined problem is also solved by a computer program product according to the invention and a computer-readable medium on which the computer program product is stored. The computer program product is designed to be executable in at least one processor. The computer program product can be stored as software, e.g., as an app downloadable from the internet, or as firmware in memory and designed to be executable by a processor or an arithmetic logic unit (ALU). Alternatively or additionally, the computer program product can also be designed, at least partially, as a hard-wired circuit, for example, as an ASIC (Application-Specific Integrated Circuit). The computer program product according to the invention comprises instructions that cause the engine control unit according to the invention to execute the process steps of the described method.The computer program is thus configured to carry out the method for operating a three-phase motor using a motor control unit. In particular, it is configured to consider a rotor flux-related switching criterion in addition to a grid-related switching criterion. Furthermore, the computer program includes commands that cause the switching signal unit of the motor control unit according to the invention to generate one or more switching signals for one or more power electronic actuators if a decision has been made to switch on a particular actuator. According to the invention, the computer program is configured to implement and carry out at least one embodiment of the outlined method. The computer program can integrate all sub-functions of the method and thus be monolithic.Alternatively, the computer program product can also be segmented, distributing sub-functions across segments that run on separate hardware. For example, the computer program product can be partially executable within the engine control unit's control unit and partially within an external control unit. Furthermore, part of the process can be carried out within the engine control unit, and another part in a higher-level control unit, such as a PLC (Programmable Logic Controller), a handheld parameterization device, or a computer cloud.

[0021] 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 show, in schematic representation: Fig. 1 an engine control unit topology; Fig. 2 a conventional structure for generating control pulses; Fig. 3 a conventional structure for generating control pulses according to a model predictive method; Fig. 4 an example of a space vector representation; Fig. 5 a representation of a pulsating current vector after a 2-phase switching of phases a and b in the α-β coordinate system; Fig. 6 a representation of a rotation operation between a space vector diagram in an α-β coordinate system and a space vector diagram in a dq coordinate system; Fig. 7 a representation of the definition of the switching angle α; Fig. 8 a representation of the definition of the switching angle range [α min ; α max ]; Fig. 9 an illustration of the problem of zero crossing detection; Fig. 10 a space vector diagram with rotor flux vector and areas with positive and negative torque; Fig. 11 a space vector diagram with rotor flux vector, voltage vector and a range with possible switching points; Fig. 12 a space vector diagram with voltage vector and switching window; Fig. 13 a space vector diagram with rotor flux vector, voltage vector and switching window; Fig. 14 a space vector diagram with rotor flux vector, voltage vector and areas which strengthen or weaken the rotor flux linkage; Fig. 15 a space vector diagram which contains the information from Fig. 10 and Fig. 14 summarized in one presentation; Fig. 16 a space vector diagram with rotor flux vector and a flux window; Fig. 17 a space vector diagram with rotor flux vector, switching window and flux window; Fig. 18 a space vector diagram with rotor flux vector, voltage vector, switching window and flux window; Fig. 19 to 24 a time series of a space vector diagram over a period of 3.9 ms with rotor flux vector, voltage vector and rotating stator current vector; Fig. 25 to 28 a time series of a space vector diagram over a period of 6 ms with rotor flux vector, voltage vector and rotating stator current vector in the case of a rotating rotor; Fig. 29 a space vector diagram with rotor flux vector, voltage vector, switching window and flux window; Fig. 30 a space vector diagram with rotor flux vector, voltage vector, switching window and a flux window which is opposite the flux window in Fig. 29 by an angle of rotation δ rot was filmed; Fig. 31 a representation of the definition of t delay ; Fig. 32 a space vector diagram with rotor flux vector and voltage vector, which represents the limits of the switching angle range and the rotor flux angle range; Fig. 33 a flowchart according to an embodiment of the method according to the invention; Fig. 34 a structure for generating control pulses according to the invention; and Fig. 35 a flowchart of a switching process.

[0022] Fig. Figure 1 shows a motor starter topology with a motor starter 1 connected between a three-phase electrical network 5 and a three-phase asynchronous motor (= ASM) 4 acting as a load for starting and operating the ASM 4. Three current paths 76.1, 76.2, 76.3 run through the motor starter 1, each for one of the three phases of the network 5. The motor starter 1 has a network-side connection device 72 for each current path 76.1, 76.2, 76.3, with which the current paths 76.1, 76.2, 76.3 can each be electrically connected to network-side phase conductors A, B, C. The motor starter 1 has a load-side connection device 73 for each current path 76.1, 76.2, 76.3, with which the current paths 76.1, 76.2, 76.3 can each be electrically connected to load-side phase conductors a, b, c.

[0023] The motor starter 1 has a power electronic actuator 3, e.g., a pair of anti-series connected transistors (e.g., MOSFET, IGBT, etc.), for each current path 76.1, 76.2, 76.3. A voltage sensor 75 is arranged at each of the power electronic actuators 3, with which the voltage drop across the power electronic actuator 3 can be measured. A current ia, ib, ic in the respective current paths 76.1, 76.2, 76.3 is measured on the mains side, with respect to the power electronic actuators 3, using current measuring devices 74. The current measuring devices 74 can alternatively or additionally also be arranged on the load side, with respect to the power electronic actuators 3, or be integrated into the power electronic actuators 3.Additionally, the motor starter uses line-to-line measurements u_A,B, u_B,C, u_C,A on the mains side to determine the mains voltages u_A, u_B, u_C, and line-to-line measurements u_a,b, u_b,c, u_c,a on the load side to determine the motor voltages u_a, u_b, u_c. These voltage measurement methods between the mains phases A, B, C, a, b, c are represented by the symbols shown in the diagram. Fig. 1 Voltage sensors 75 are shown, with which voltages between the current paths 76.1, 76.2, 76.3 can be measured.

[0024] The motor starter 1 has a mechanical interrupter contact 70 for each current path 76.1, 76.2, 76.3, which is arranged on the mains side, viewed with respect to the power electronic actuators 3. The mechanical interrupter contacts 70 can be designed as manually operated switches 71 with a latching position and locking mechanism. Alternatively, the mechanical interrupter contacts 70 can be designed as remotely operated switches, e.g., as relays.

[0025] In each of the current paths 76.1, 76.2, 76.3, a mechanical safety switch 20, e.g., a manually operated switch, is arranged on the load side with respect to the power electronic actuators 3, with which the respective current path can be interrupted (galvanic isolation). In a first current path 76.1 of the motor starter 1 and in a second current path 76.2 of the motor starter 1, circuits are also provided with which switching between a direct circuit 20a of the motor starter 1 (forward operation of the ASM 4) and a reversing circuit 20b of the motor starter 1 (reverse operation of the ASM 4) is possible.

[0026] Although the range of functions of engine control units has been continuously expanded in recent decades, e.g. through voltage ramping, current-limited operation, torque-controlled ramp-up and speed-controlled ramp-up, the structure for generating control pulses has remained essentially the same and has not changed in the Fig. The structure shown in point 2 essentially comprises three blocks, which are explained below: In a first block of the engine control unit 1, the controller 21, the setpoint of the ignition angle α, also called control or trigger angle, is determined by a control structure from measured values ​​24 such as stator current I. 1,RMS , Stator voltage U 1,RMSand / or rotor speed n of the motor. The ignition angle α can be influenced by input parameter 25, e.g., the current motor torque, the maximum set current, or a speed ramp, to optimize the behavior for the application. The ignition angle α defines the RMS value of the voltages applied to the individual phases a, b, c of the motor (RMS = Root Mean Square). A simple way to generate an ignition angle α is a voltage ramp, independent of the load response; for this, the ignition angle α is simply increased as a function of time. A sufficiently large initial value of the ignition angle α must be used to ensure that the initial motor torque M M already greater than the load torque M L is; otherwise, the motor current would not accelerate the rotor and would only generate high losses at standstill until the motor torque is high enough to accelerate the rotor.

[0027] A second block of the engine control unit 1, a control signal block 22, serves to generate control signals 27 for triggering ignitions for the thyristors, wherein the control signals 27 are generated depending on a measured voltage, e.g. the mains voltage 26, and / or a measured current, preferably with a defined time between a thyristor being switched off when a minimum current is undershot and the thyristors being re-ignited.

[0028] A third block of the engine control unit 1, a hardware block 23, comprises the thyristors and ignition devices, e.g. an ignition signal unit; it uses the control signals 27 received from the control signal block 22 to generate the output voltage.

[0029] Fig. Figure 3 shows a structure for generating control pulses according to a model predictive control method, as described in the aforementioned patent application EP 3916998 A1 and in Nannen et al. The engine control unit 1 can, for example, be implemented in a topology such as that described in Fig. The engine control unit 1, as shown in Figure 1, is used. The engine control unit 1 has a control unit 41 with a processing unit 46, also referred to as a processor, and a memory unit 43. The processing unit 46 executes a computer program stored in the memory unit 43, which contains an algorithm for carrying out the procedure. During execution of the algorithm, ignition possibilities for at least one subsequent time step are determined, the corresponding engine behavior is predicted using an engine model, and finally, based on the predicted engine behavior, a decision is made as to whether and which ignition possibility should be selected. To initialize the algorithm, state variables of the system are measured or calculated. The processing unit 46 receives a series of measured values ​​44, e.g., the engine current I1, the engine voltage U1, and the rotor speed n, as input values. Mains voltages u A , u B , uC are measured to the grid angle φ grid and the mains voltage amplitude U grid to calculate. The motor currents are also calculated. a , i b , i c and motor voltages u a , u b , u c Used for a model-based calculation of the rotor flux. Together with the measured mechanical rotor speed n and the grid frequency f. gridAll system variables are defined. After prediction, it is determined which ignition possibilities exist at the considered future time, i.e., which ignition possibilities exist. These ignition possibilities can then be checked and evaluated against a variety of decision criteria 45, e.g., a maximum torque or a maximum stator current. If a decision has been made for a specific ignition possibility, the controller 41 generates one or more control signals 47 to an ignition block 42 of the engine control unit 1, which contains the thyristors and an ignition signal unit 48 for generating ignition signals for the thyristors. The control signals 47 cause the ignition signal unit 48 to generate ignition signals for one or more of the thyristors, so that the pre-calculated engine behavior is achieved. Basis - Space vector representation

[0030] The present description of the invention is based on space vector representations: these offer the possibility of representing multiphase physical quantities occurring in a multiphase network in a simpler, two-axis coordinate system with axes α and β as space vectors rotating around the origin, or simply: vectors. The transition to the α-β coordinate system is achieved by applying the Clarke transform. Using space vectors, physical quantities in the system can be represented at any point in time; in the present invention, these are primarily magnetic flux, voltage, and current.

[0031] Fig. Figure 4 shows an example of a space vector representation: In the upper part of Fig. Figure 4 shows a typical graph of the waveform of a three-phase AC voltage U and a stator current I, also known as motor current, over time. The three-phase AC voltage U originates from a typical 50 Hz power supply network and consists of three individual sinusoidal AC voltages u. A , u B , u C of the same frequency with a peak value of 325 volts, whose phase angles are fixed at 120° relative to each other. The voltage waveform is shown over the period from 0 to 100 ms. The motor current i a , i b , i c In the three motor phases a, b, c, shown over the period from 0 to 30 ms, a single two-phase switch-on in phases a and b at time 12 ms in a configuration as shown in Fig. 1, generated with an engine control unit and a connected engine.

[0032] The lower part of Fig. Figure 4 shows space vector representations of voltage and current in the time interval from 12 ms to 21 ms, in time steps of 1 ms. The scaling of the vectors is irrelevant for the analysis here. The voltage vector is represented by a thicker line. U→ The current pointer is shown with a thinner line thickness. I→. At time t = 12 ms, no current is flowing; therefore, only the voltage phasor is present. U→ to see. When comparing the images at times t = 12 ms and t = 13 ms, it is immediately noticeable that the voltage phasor U→ It has rotated a little further. This can also be seen in the subsequent time steps. According to the three-phase alternating voltage, this is a rotating voltage phasor. U→, proportional to the mains frequency f grid = rotates at 50 Hz: the voltage pointer U→ rotates with an angular frequency ω = 2πf grid= 2π 50 Hz = 2π / (20 ms) counterclockwise around the zero point located in the center of each time step window. The amplitude, and thus the length of the voltage pointer, is determined by the voltage pointer. U→ They remain constant in an ideal network where there is no impedance and no voltage drops on the line. Furthermore, the space vector representation at time t = 13 ms shows that the onset of current flow forms a current phasor. I→ This leads to a current phasor that initially shows a small amount. From the space vector representations for the subsequent time steps, it can be seen that the current phasor... I→ its length continuously increases while its direction remains constant; this is typical for two-phase ignitions. At time t = 16 ms, the current phasor has I→ Once it reaches its maximum, its length decreases in each time step until the current flow ends at time t = 21 ms.

[0033] This example shows how the space vector representation can be used to clearly illustrate relationships between different physical quantities, in this case voltage and current. Basis - Two-phase ignitions

[0034] The engine control unit allows for various switching operations, which trigger different engine behaviors. Switching is either two-phase (phases a&b, b&c, c&a) or three-phase (phases a&b&c). The two-phase switching, as already mentioned in Fig. Figure 4 shows current vectors that are fixed in their orientation and pulsating in their amplitude.

[0035] When the power electronic actuators are switched on in the two mains phases a, b of the three mains phases a, b, c, these phases a, b become conductive, whereas the power electronic actuator in the third phase c remains inactive. Because the sum of the currents I in all three phases a, b, c must always be zero, the current flowing towards the motor in one of the conductive phases a, b must be identical to the current flowing away from the motor in the other conductive phase; therefore, the current amplitudes in the two conductive phases a, b are exactly the same, only with opposite signs.

[0036] Fig. 5 concerns a two-phase connection of phases a and b. In the left diagram of Fig. 5. In a fixed α-β coordinate system, it is obvious that this condition leads to a pulsating current vector. la→+lb→ along the railway line I c =0 leads to the current vector la→+lb→ Although its length varies, its direction remains constant. Therefore, only the absolute value |i varies. a +i b | of the current vector la→+lb→. With 2-phase switching, fixed paths 7 result on which pulsating current space vectors can be generated; in the left diagram of Fig. These 5 lanes are 7 with I a = 0, I b = 0 and I c = marked 0, where the track I a = 0 runs along the β-axis and the path I b = 0 forms an angle of 30° with the α-axis; each of the paths 7 forms an angle of 60° with the nearest adjacent path 7. In a three-phase connection, however, the current also forms a rotating space vector – similar to the voltage. In the right-hand diagram of Fig. 5 are the current amplitudes i a , i b in the two conducting phases a, b as well as the absolute value |i a |+|i b| of the current vector generated by a 2-phase ignition la→+lb→ represented over time t. Fundamentals - Engine modeling and rotary operations

[0037] The present invention is generally applicable to a three-phase motor connected to a motor control unit. The three-phase motor can be designed as an asynchronous motor, a synchronous motor, or a PM line-start motor.

[0038] An asynchronous motor model was explained in detail in the European patent application EP 3 916 998 A1 mentioned in the introduction to the description, as well as in Nannen et al. All definitions of symbols given there are also used in the present description. The motor model equations described in these publications, which are used here as a basis, are: dI→1Sdt=1σ⋅L1⋅U→1S−R1⋅L22−R2⋅Lh2σ⋅L1⋅L22⋅I→1S+ R2⋅Lhσ⋅L1⋅L22⋅Ψ→2S−j⋅ΩL⋅Lhσ⋅L1⋅L2⋅Ψ→2SdΨ→2Sdt=Rs⋅LhL2⋅I→1S−R2L2⋅Ψ→2S+j⋅ΩL⋅Ψ→2S MM=32⋅p⋅LhL2⋅Ψ→2S×I→1S

[0039] The subscript 1 identifies a stator-related quantity, the subscript 2 a rotor-related quantity. The superscript S indicates that the quantity is referenced to a stator-fixed coordinate system. The motor model uses the stator current. I→1S, the scattering coefficient α (Blondel's coefficient), the inductances L1, L2, the voltage U→1S, the resistors R1, R2, the linked rotor flux Ψ→2S, the main inductance L h (English: mutual inductance), the rotor speed Ω L and the engine torque M M .

[0040] In the aforementioned European patent application EP 3 916 998 A1 and in Nannen et al., a coordinate transformation is introduced at this point, which changes the coordinates in the α-β coordinate system by the rotation angle φ. K turns: I→1K=I→1S⋅e−jφκ

[0041] The angle of rotation φ K is chosen such that the rotor flux vector Ψ→2S After the rotation is completed, the vector points in the positive direction of the d-axis of the dq coordinate system. The value d represents the magnetic flux density of the magnetic excitation in the rotor, and q is an expression for the torque generated by the rotor. To clarify: All vectors are now rotated by the angle φ in the α-β coordinate system. K rotated. Therefore, the current is: I→1K=I1d+j I1q

[0042] This rotation operation forms the basis for the entire theory of field-oriented control of motors at the inverter and is explained here to make the processes under consideration easier to understand. In this context, it is referred to as rotor flux orientation. In equations (3) and (4), the index K indicates that the indexed quantity refers to an arbitrary coordinate system.

[0043] Fig. Figure 6 illustrates the rotation operation: in the left part of the Fig. 6 is the position of the flow indicator Ψ→2S before rotation by the angle of rotation φ K shown and the rotation of the flow pointer by the rotation angle φ K As indicated above with reference to Fig. As explained in section 5, a two-phase connection results in a current vector with a fixed direction but pulsating amplitude. These fixed directions of the current vector, which arise from the three possible two-phase connection combinations a&b, b&c, and a&c, are described in Fig. 6 dashed line 7 drawn; in this description these fixed directions 7, in which the current vector generated by a 2-phase connection pulsates, are also referred to as paths.

[0044] This rotation operation 6 will change the flow indicator Ψ→2S and lanes 7 into the right part of the Fig. The 6 positions shown have been brought to light: the river pointer Ψ→2S points in the direction of the d-axis.

[0045] In the dq coordinate system, one can distinguish a first region 8 with additions for positive torque, comprising the dashed-bordered half-space with positive q-values, and a second region 9 with additions for negative torque, comprising the dashed-bordered half-space with negative q-values, which border each other along the d-axis.

[0046] The representation obtained by the rotation operation 6, in which the dq coordinate system is at each time at the rotor flux pointer Ψ→2S If the rotor is aligned, meaning the entire rotor flux lies in the d-direction, it is possible to divide the current into a field-forming current, which corresponds to the current phasor component along the d-axis, and a torque-forming current, which corresponds to the current phasor component along the q-axis.

[0047] One of the major advantages of the representation in the dq coordinate system is that equation (2) for the motor torque M M greatly simplified and becomes a simple multiplication: MM=32⋅p⋅LhL2⋅Ψ2d⋅I1q

[0048] Ψ 2d is the d-component of the linked rotor flux in the dq coordinate system, I 1qThe q-component of the stator current in the dq coordinate system: Only the components of rotor flux and stator current that are perpendicular to each other contribute to the motor torque. Basis - Conversion of classical methods with switching angles into vector representation

[0049] The methods commonly used in industry today for controlling engine control units are based on an ignition angle α. This can be related to the mains voltage or the phase current. The currently used algorithm serves as the starting point for deriving the switching algorithm according to the invention, which requires relatively little computing power. Fig. Figure 7 illustrates the conventional definition of the firing angle α, which determines the firing point in relation to the mains voltage, more precisely: to the zero crossings 10 of the mains voltage. A firing of the thyristors therefore always has a fixed interval α from the immediately preceding zero crossing 10 of the mains voltage. Fig. Figure 7 also shows the current pulses i(t) triggered by the respective ignitions as dashed curves; the sign of a current pulse i(t) corresponds to the sign of the voltage u(t) at the respective ignition time.

[0050] According to the present invention, the firing angles of thyristors are now replaced by switching angles of power electronic actuators.

[0051] The conventional definition of the ignition angle α, which in the present invention now appears as the switching angle α, complicates the integration of further criteria regarding a switching decision: since a precise switching time is defined at which the switching must occur, a further criterion could only influence whether the switching takes place at all or not. Since this restrictive definition leaves no room for including a further criterion, e.g., the rotor flux as an additional switching criterion besides the grid angle / grid voltage zero crossing, this definition is modified as shown in Fig. Figure 8 shows that instead of a specific switching-on time, the switching-on angle α, an angular range, the so-called switching-on angle range 11, is now defined within which the switching-on time should lie. The switching-on angle range 11 is also referred to as the switching-on sector or turn-on window. The switching-on time can be selected from a range of values ​​[α min ; α max ] are chosen, which is defined by a lower limit α min , the minimum ignition angle, and an upper limit α max, limited by the maximum ignition angle. The introduction of the switching angle range 11 allows for more alternatives in choosing a switching point, since not only a specific point in time, but also an angular range and thus a time interval is available for switching. Compared to the reference point "zero voltage crossing" of the conventional switching angle α, the reference point of the switching window is shifted by 90 degrees. Of course, configurations are also possible in which the width of the switching window 11 is zero (α). min = α max ), meaning that only a single activation time is possible.

[0052] Due to the introduction of a switching window, it is not just a single current pulse, as in Fig. 7 represented as i(t), possible, but a set of current pulses 120, as in Fig. 8 drawn as dashed curves. How Fig. Figure 8 shows that choosing the smallest angle α results in minA current pulse of 120 with the largest possible current amplitude. The larger the switching angle α is chosen, i.e., the closer the switching angle is to the upper limit α. max The closer the angle is, the smaller the current amplitude of the resulting current pulse becomes; 120 when choosing the largest angle α maxThis results in a current pulse 120 with the smallest possible current amplitude. A basic consideration of the asynchronous motor as a load explains this relationship: in a very simplified way, an asynchronous motor can be represented as a series circuit of an inductor and a resistor; it therefore always behaves in a resistive-inductive manner. This means that the current flowing due to an applied voltage always lags behind this voltage in time. For the generation of the control pulses, this means that the connection must always be made before the corresponding voltage zero crossing 10. The earlier the connection is made before the voltage zero crossing 10, the greater the resulting current.

[0053] By choosing the upper limit α maxThe switching angle range 11 ensures that even if the voltage zero crossing detection does not accurately detect the zero crossing, no unexpected switching pulse is generated; such unexpected switching pulses can damage the engine control unit-engine system due to their very high current amplitudes. Fig. Figure 9 illustrates, in an example, the consequences that can occur if a zero crossing of the mains voltage is not correctly detected: the amplitude of a sinusoidal mains voltage u(t) is plotted against time t. The power electronic actuators are switched on at a fixed angular interval α from the zero crossings 10 of the mains voltage u(t). Due to high harmonic distortion or other distortions in the region before and after the zero crossing 10 of the mains voltage, a non-sinusoidal voltage range 51 can arise. The theoretical zero crossing 10' of the ideal mains voltage u'(t) is delayed by a delay period 52 due to the non-sinusoidal voltage range 51; accordingly, the measurement of the switch-on angle α does not begin at the theoretical zero crossing 10', but at the actually measured zero crossing 10.The subsequent switching Z therefore occurs, due to the delay of the last voltage zero crossing 10, not before, but after the next voltage zero crossing 10: thus, instead of a very large control angle, a very small control angle is set, and a second current pulse i2(t) is generated whose amplitude is many times greater than that of the preceding first current pulse i1(t). Such an unfavorable scenario can be avoided by appropriately choosing the upper limit α. max of the switching angle range 11 are excluded, so that only those switching pulses are allowed through which lead to acceptable current amplitudes, even if there are uncertainties in the system regarding the zero crossing of the mains voltage. Flow-oriented start algorithm for motor control unit-controlled induction motors

[0054] The preceding principles and relationships form the basis for explaining the invention. The following conclusion can be drawn from equation (5): Does the stator current I 1q A positive value results in a positive torque. A negative sign results in a negative torque. This can be used to evaluate switching options.

[0055] For further derivation, it is now assumed for simplification that a rotor flux Ψ2 with such a large d-component Ψ 2d The condition is that a torque is generated (see equation 5) which can set the motor into a measurable rotational motion, but the rotor does not rotate. Furthermore, the paths 7 of the theoretically possible two-phase switching operations in the α-β coordinate system are known. Therefore, it can be precisely determined which switching operation can generate positive torque; see the dashed-bordered half-space 8 in Fig. 10, and with which switching negative torque can be generated, see the dashed-bordered half-space 9 in Fig. 10.

[0056] However, in the space vector representation, the network is described by a voltage vector that rotates at the angular frequency of the mains voltage. Therefore, not all switching possibilities are feasible, but only those that lie within the theoretically possible range, which in the space vector representation is the half-space from +90° to -90° around the voltage vector. See illustration. Fig. 11: it contains the space vector diagram of Fig. 10 around an example voltage pointer U→1S and added a section 13, which contains the possible connections. Fig. 11 is compared to Fig. 10. Area 13 has been added, in which those switching configurations (railways) are located that are possible due to the network location.

[0057] By including the voltage indicator U→1S All switching possibilities that are physically possible at this point in time have now been shown, i.e., those switching possibilities that lie within range 13. However, this does not clarify the maximum amplitude of the resulting current. To define a limiting measure for the maximum current, a definition similar to that of the ignition angle α in the conventional method is used. However, instead of defining a fixed ignition angle α, here: switching angle, as in the conventional method, a variable value is now used, as already mentioned in connection with... Fig. 8 explains the switching angle range 11 with a minimum switching angle α min and a maximum switching angle α max defined, which are in a fixed angular relationship to the stress vector U→1S are located and move with it at the angular velocity ω u = 2πf grid, which are directly proportional to the mains frequency f grid is, rotate, see Fig. 12. To illustrate: In Fig. 12. The switching options would be those whose paths lie within the switching angle range 11, i.e., only the switching that corresponds to path i. b = -i c This corresponds to the addition of phases b and c.

[0058] After the range of permitted switching-ins is limited to the switching angle range 11, bounded by the minimum and maximum switching angles α min and α max Since the range has been restricted, the torque range should also be restricted. If the goal is to accelerate the engine, all activation options in the first range 8 are attractive because they generate positive torque, and all options in the second range 9 are unattractive because they generate negative torque.

[0059] In addition to the torque criterion, it must also be ensured during the operation of the three-phase motor that sufficient flux remains in the motor after an additional current is applied. For example, if an additional current were applied along the path i... a = -i c in Fig. If this occurs in step 11, positive torque would be generated, since this activation takes place in the half-space 8, in which positive torque +M M can be generated, is located. However, the one along the path i a = -i c pulsating current pointers I→1S largely opposite to the river indicator Ψ→2S This is oriented and would therefore lead to a significant reduction in flux within the rotor. This analysis is based, for simplicity, on a single specific activation; in reality, it is the superposition of several individual activations that statistically leads to a reduction in flux.

[0060] Fig. Figure 14 shows, neglecting rotor resistance R2 and iron losses, where flux would be built up and where it would be reduced: there is a dashed-bordered area 14 in which the switching options are located that strengthen the rotor flux linkage, and a dash-dotted-bordered area 15 in which the switching options are located that weaken the rotor flux linkage.

[0061] Fig. 15 summarizes the information from Fig. 10 and Fig. 14 in one representation: areas 8 and 9, where positive torque +M M or negative torque -M M is generated, and areas 14 and 15, where the river is built up Ψ→2S⇑ or river Ψ→2S⇓ These overlap and lead to the four different sectors 16 to 19 shown. The in Fig. The 15 plotted areas 8, 9, 14, 15, 16, 17, 18 and 19 are in a fixed angular relationship to the rotor flux indicator. Ψ→2S and rotate with it at the angular velocity ω Ψ = 2πnp, which is directly proportional to the mechanical rotor speed n and the number of pole pairs p of the three-phase motor.

[0062] If the engine control unit's goal is to accelerate the engine, the selected current pointers should be located in the sectors with positive torque, i.e., sectors 16 and 19. Due to the constant decrease in rotor flux caused by the rotor resistance R2, it is beneficial if, on average, the switching operations generate more flux than is lost. Fig. Figure 16 shows a flux and torque angle range 12, also referred to simply as the flux window, which takes this behavior into account: the flux and torque angle range 12, which relates to the rotor flux pointer Ψ→2S refers to the angle γ start and γ end limited, which is determined by the rotor flux indicator Ψ→2S from which measurements can be taken. The flux and torque angle range 12 was determined based on the relationships in Fig. 15, together with the requirement that, on average, the ignitions should generate more flux than is lost, is specified as an example.

[0063] The now defined flux and torque angle range 12 therefore includes sectors with positive torque and sectors with flux build-up or only slight flux reduction. The two defined ranges, the activation angle range 11 and the flux and torque angle range 12, are now to be combined into a single decision criterion. If an activation option 7 lies within both the activation angle range 11 and the flux and torque angle range 12, this activation should be carried out. In the Fig. In example 17, this is not the case for any of the switching options; therefore, no switching should be carried out. Now, one can wait a certain amount of time, during which the voltage pointer changes. U→1S continues to rotate. For example, approximately 5 ms later, which corresponds to a 90° rotation in the space vector diagram for a 50 Hz AC voltage with an angular frequency of ω = 2π / (20 ms), the constellation looks like this: Fig. 18: The voltage indicator U→1S The rotor flow indicator has rotated another 90°. Ψ→2S In contrast, its amplitude and direction remained constant (motor is stationary, no iron and rotor losses). Now, path 7 of the switching combination b&c lies in both sector 11 and sector 12. Therefore, this switching combination b&c can be executed to generate positive torque +M M to generate the rotor flux Ψ→2S to increase the size of the current without exceeding a maximum current amplitude.

[0064] The procedure described here can now be executed repeatedly with a fixed step size. Whenever one of the switching combinations a&b, b&c, c&a lies within sectors 11 and 12, it is activated. Thus, a simple criteria check is sufficient to generate the control pulses.

[0065] Once a connection has been established, it is possible that, while current is still flowing, another connection opportunity arises involving the third, still-blocking power electronic actuator. In this case, this connection would also be executed immediately. Accordingly, a three-phase connection a&b&c occurs, which, unlike the two-phase connections, results in a rotating current phasor. I→1S This situation of a 3-phase connection is found in the Fig. 19 to 24 illustrated. Fig. 23 at t = 3.4 ms and Fig. 24 at t = 3.9 ms show that when the current phasor I→1S The next switching point occurs during the movement, and it then "locks" into that position. Considering the associated phase currents, this is the point at which the power electronic actuator blocks in one of the phases due to a current zero crossing. Rotating engine

[0066] All previous considerations are based on the simplifying assumption that the rotor is stationary. However, this assumption is no longer valid after the first switching operations, once the rotor has started moving. Therefore, the motor's rotational speed must now be taken into account. The fixed paths 7 of the current phasor are shown in the α-β diagram in Fig. The 5 values ​​are unaffected by the rotational speed, as they depend solely on the stator, which does not rotate. The fundamental relationships of the switching angles also remain unchanged. However, the position of the rotor flux vectors in the space vector representation changes during current flow. Ψ→2S and the stator current pointer I→1S. More precisely: It is not only the position of the space pointers that changes. Ψ→2S and I→1S, But the angular position of these two space vectors relative to each other also changes. This deviation leads to a deviation in the assessment of the expected torque and its influence on the rotor flux. These are discussed in Fig. 25 to 28 are illustrated by way of example.

[0067] Fig. 25 to 28 apply to an activation for a machine with a mechanical rotational speed of n = 250 min⁻¹ -1 (Rated speed 1470 min) -1 ) rotating rotor. The voltage indicator U→1S rotates with a frequency f relative to the mains frequency grid directly proportional angular velocity ω U = 2πf grid , the river pointer Ψ→2S with an angular velocity ω directly proportional to the number of pole pairs p and the mechanical rotor speed n Ψ = 2πnp, where these two angular velocities do not coincide; in contrast, the current phasors 7, the so-called paths, point in fixed directions for 2-phase ignitions. In Fig. From 25 to 28 it is visible that, due to the rotation of the rotor, next to the voltage indicator U→1S also the river indicator Ψ→2S rotates. Accordingly, the fixed relationship between the current pointer applies. I→1S At the beginning of the engagement, the current flow is no longer constant throughout the entire duration of the current flow, but rather changes over time. This means that considerations regarding positive torque and rotor flux development are based on a different assumption than what actually occurs during the engagement. To compensate for this effect, the originally defined range for possible engagements is adjusted by a twist angle δ. rot twisted. This twisting is intended to compensate as best as possible for the deviation caused by the rotating rotor.

[0068] Determining the angle of rotation δ rotThis could look like the following, although other systematic determinations are also conceivable: assuming that the conditions at the midpoint of a current pulse approximately represent the average of the conditions over the entire current pulse, this can be used as a reference. Now, the last current pulse is used as a basis, and the time interval t is determined here. delay between the switching-on time and the current maximum, as in Fig. Figure 30 illustrates this. It can now be assumed with a good approximation that the time interval t delay In the case now to be evaluated, it is identical to that of the previous switching pulse. In other words: If, in the previous switching pulse, the time interval between switching on and the current maximum is equal to t delay If the current value is present, it will likely be very similar or identical at the next switching pulse. Accordingly, the flux and torque angle range 12 can be adjusted by a factor corresponding to the time interval t. delaycorresponding angle of rotation δ rot pre-filmed, as in Fig. 30 shown to minimize the influence of the rotating rotor on the switching decision.

[0069] The following equation (6) can be used to determine the angle of rotation δ. rot in this approach, where Ω L The rotor speed is: δrot=2π(tdelay1ΩL)=2πΩLtdelay

[0070] Will the angle of rotation δ rot By determining the parameters in the manner described, a large proportion of the changes caused by the rotating rotor can be compensated. Influencing behavior by influencing the position of the switching limits

[0071] Up to now, the definition of the two switching ranges has been very vague and based on principled considerations to explain the underlying concept. In practical application, however, a systematic approach is needed to define these values ​​so that the actual operating behavior corresponds as closely as possible to the desired operating behavior. For this to happen, the limits must be systematically determined. Fig. 32 are the limits α min , α max , γ start , γ end The process is presented, and its influence on the procedure is systematically explained below. Upper switching angle range limit α max :

[0072] This boundary is of crucial importance; it indirectly defines the shortest current flow through a switching point. This angle α maxIt should be chosen so that even in a network where zero-crossing determination is difficult (high harmonic load or distortions in the area before and after the zero crossing), it is ensured that the switching always behaves in such a way that the current flows in the desired direction and that there is no unwanted switching with much too high a current amplitude. Lower switching angle range limit α min :

[0073] This limit determines the switching point at which the current flow reaches its maximum length and greatest amplitude. This angle α min Analogous to the ignition angle α in classical control methods, it can be specified, for example, via a characteristic curve or controlled taking the current amplitude into account. Considering the current flux amplitude to limit the maximum possible torque is also conceivable. Since the angle α minSince the current and torque can be controlled, the function of this angle is similar to that of the ignition angle α in conventional control methods. min can take a value from the following domain: [-90°, a max ], measured in relation to the voltage pointer U→1S. Limit to negative torque γ start :

[0074] This limit defines the boundary with negative torque. Depending on the choice of angle γ. start For example, a slightly negative torque is allowed in favor of a strong flux-forming current. Flow control limit γ end

[0075] This angle γ end determines how long the switching operations continue to generate positive torque +M Mto generate. Depending on how this limit is defined, either the system is switched on for a long time, thereby reducing the rotor flux linkage, or it is only switched on if, on average, there is no reduction in the rotor flux linkage. By defining the angle γ end The amount of rotor flux linkage remaining in the rotor and present during the next ignition phase is actively determined. This parameter can be used for "flux guidance" or "flux control" during startup. This is achieved using the angle γ. end can therefore the mean absolute rotor flux |Ψ→2S| can be influenced. If the engine is to be accelerated, the difference γ can be end - γ start assume a value from the following range of definitions: [90°, 180°], measured with respect to the flux indicator Ψ→2S. Exemplary rule structure for influencing switching limits

[0076] As already explained in the previous section “Influencing the behavior by influencing the position of the switching limits”, the angles α can min , α max , γ start , γ end , which limit the switching windows or flow windows, are used to influence the ramp-up.

[0077] The invention also enables controlled operation. The preceding description of the relationships above assumes that the goal is to accelerate the rotor shaft; this requires the generation of positive torque. The same arguments can also be applied to braking, except that here the range of possible activations is located in the area for negative torques. Transition between processes - combination with conventional processes

[0078] The method presented here offers advantages over conventional control methods based solely on a switching angle α, but presumably also some disadvantages. Therefore, the invention can be used not only exclusively for determining control signals, but also as a complement to existing methods. It is possible that the present method is better suited for certain speed ranges than for others. Accordingly, it is also possible to combine the new method presented here with a known method during startup. In a preferred embodiment, the startup could initially begin with the new method; subsequently, during startup, a decision is made based on a quality criterion, e.g., the speed, to switch to a conventional control method in order to combine the advantages of both methods. Rotor flux as a basis

[0079] All assessments are based on the use of the linked rotor flux. Ψ→2S. In contrast to the stator current, the linked rotor flu Ψ→2S However, in an industrial environment, it cannot be directly measured using appropriate sensors. The simplest way to determine the rotor flux linkage is to use the modeling method introduced above (see equation (16) in Nannen et al.): dΨ→2Sdt=R2⋅LhL2⋅I→1S−R2L2⋅Ψ→2S+j⋅ΩL⋅Ψ→2S

[0080] For this purpose, the stator current is used. I→1S and the rotor speed Ω L required. The stator current I→1S It can be easily measured: the two or three phase currents are measured, then the current vector is determined using the Clarke transformation. I→1S The mechanical speed can, in principle, also be determined using suitable measurement technology. However, this is atypical when using motor control units in industrial environments, as the costs and effort of installation make their use uneconomical. Accordingly, determining the speed without an additional sensor appears very advantageous. For determining the speed based on models, there are numerous possibilities that have been developed in recent decades with regard to the sensorless control of three-phase machines. An overview is provided by the following monograph: Vas, Peter: Sensorless Vector and Direct Torque Control. Monographs in Electrical and Electronic Engineering, Volume 42. Oxford: Oxford University Press 1998, ISBN-13: 978-0198564652. An alternative modeling of the rotor flux is also possible.

[0081] Fig. Figure 33 shows a flowchart according to an embodiment of the method according to the invention. In a first step S1, the grid-related switching criterion, the rotor-related switching criterion, and the switching-off criterion are defined. The grid-related switching criterion is fulfilled if a switching possibility lies within the switching angle range 11, which has a lower switching angle range limit α. min and an upper limit of the switching angle range α max with regard to the voltage indicator U→1S. The rotor-related switching criterion is fulfilled if a switching possibility lies within the flux and torque angle range 12, whose range limits γ start and γ endare defined with respect to the rotor flux. The grid-related shutdown criterion with respect to a power electronic actuator 3, which can switch a grid phase a, b, c, is fulfilled when a current zero crossing occurs in this grid phase a, b, c. Since only a grid-related shutdown criterion exists, but no rotor-related shutdown criterion, the grid-related shutdown criterion can simply be referred to as the "shutdown criterion" for the sake of simplicity. In the first step S1, power electronic actuators are also switched on to set the rotor in motion so that the rotor flux is not zero.

[0082] In a second step S2 following the first step S1, it is checked whether the grid-related switching criterion is met. The grid-related switching criterion is met if a switching option of two or more grid phases a, b, c lies within the switching angle range 11.

[0083] In a third step, S3, following the second step S2, it is checked whether the rotor flux-related switching criterion is met. The rotor flux-related switching criterion is met if a switching possibility of two or more network phases a, b, c lies within the flux and torque angle range 12.

[0084] In a fourth step, S4, following the third step S3, the corresponding power electronic actuators are switched on if two or more network phases a, b, c can be switched on, provided that both the network-related switching criterion and the rotor flux-related switching criterion are met. After the fourth step S4, the process returns to the second step S2, from where the iteration of steps S2 to S4 begins again. Engine control unit topology

[0085] The basis for all considerations made is the one in Fig. Topology 1 shown. In principle, the method can also be applied to any other topology of power electronic actuators.

[0086] One possible structure for generating control pulses according to the present invention is in Fig. Figure 34 shows that in a first block of the motor control unit 1, a control unit 61 receives setpoints 65, e.g., a desired motor torque, a maximum setpoint current, or a speed ramp, and measured values ​​64, e.g., stator current I1, stator voltage U1, torque M, and / or rotor speed n of the motor. Based on the received setpoints 65 and the received measured values ​​64, the control unit 61 sets one or more of the actuated values ​​68, i.e., α min , α max , γ start , γ end, so that the target values ​​can be achieved. These control values ​​68 are transmitted to a second block of the engine control unit 1, a computing unit 62, which has a processor and a memory unit. The processor executes a computer program stored in the memory unit, which contains an algorithm for carrying out the method according to the invention. The computing unit 62 generates control signals 67 from the received control values ​​68.

[0087] A third block 63 of the engine control unit 1 comprises power electronic actuators and switching devices, e.g., a switching signal unit. The switching devices control the switching states (high impedance, low impedance) of the power electronic actuators using switch-on and switch-off signals, e.g., with voltages of different magnitudes. The power electronic actuators can be designed as transistors with a gate terminal, and the switching devices as gate drivers, which control the transistors by applying voltages of different magnitudes to the gate terminal. The control signals 67 received by the processing unit 62 cause the switching signal unit to generate switch-on signals for one or more of the power electronic actuators, so that the power electronic actuator(s) are switched on.

[0088] Fig.Figure 35 shows the sequence of a switching operation, including the switching off of the power electronic actuator. In a first step 81, the power electronic actuator is in a high-resistance, current-blocking state. In a second step 82, the power electronic actuator is switched on. In a third step 83, the power electronic actuator is in a low-resistance, current-conducting state. In a fourth step 84, a zero-crossing of the current is detected, and, triggered by this, the power electronic actuator is switched off in a fifth step 85 (high-resistance, current-blocking). In a subsequent step 86, in which the power electronic actuator is switched off, the switching operation loops back to the first step 81. Reference symbol list 1 Engine control unit 3 power electronic actuator 4 Asynchronous motors 5 network 6 Rotation operation 7 lane 8 area with ignitions for positive torque 9 area with ignitions for negative torque 10 Zero crossing of the mains voltage, measured 10' Zero crossing of the mains voltage, theoretically 11 Ignition angle range 12 Flux and torque angle range 13 areas with possible ignitions 14 Area with ignitions that strengthen the rotor flux chain 15 Area with ignitions that weaken the rotor flux chaining 16 angle range, sector 17 Angle range, sector 18 Angle range, sector 19 Angle range, sector 20 mechanical safety switches 20a Direct switching (forward operation) 20b Reversing circuit (reverse operation) 21 Controller 22 Control signal block 23 Hardware block 24 measured value 25 parameters 26 Mains voltage 27 Control signal 30 speed controllers 31 Decoupling unit 32 flow regulators 41 Control unit 42 Ignition block 43 storage units 44 Measured value 45 Decision criterion 46 computing units 47 Control signal 48 Ignition signal unit 51 non-sinusoidal voltage range 52 Delay duration 61 Control unit 62 computing units 63 power electronic actuators and switching devices 64 measured values 65 Target value 66 measured values 67 Control signal 68 Position value 70 Mechanical contact 71 Manual switch with detent position and locking mechanism 72 Mains-side connection device 73 Load-side connection device 74 Current sensor 75 Voltage sensor 76.1 Current path 76.2 Current path 76.3 Current path 81st process step 82nd process step 83 Procedure step 84th process step 85 Procedure step 86th process step 120 current pulse phase (mains side) A α Phase (load side) Phase B (grid side) b Phase (load side) Phase C (grid side) c Phase (last side) f grid Mains frequency I a Current in phase a I b Current in phase b I c Current in phase c i electricity i1 power i2 power n mechanical rotor speed p number of pole pairs S1 Procedure step S2 process step S3 process step S4 Procedure step t time t delaythe angle of rotation δ rot corresponding time interval and voltage Ignition timing; ignition timing α Ignition angle α min lower ignition angle range limit α max upper ignition angle range limit γ start lower flux angle range limit γ end upper flow angle range limit δ rot Angle of twist ω angular velocity Stator current indicator Mains voltage indicator Rotor flux indicator QUOTES INCLUDED IN THE DESCRIPTION

[0000] This list of documents cited by the applicant was automatically generated and is included solely for the reader's convenience. The list is not part of the German patent or utility model application. The DPMA accepts no liability for any errors or omissions. Cited patent literature

[0000] EP 3916998 A1 [0002, 0029, 0038, 0040] Zitierte Nicht-Patentliteratur

[0000] Nannen et al.“ referenzierten Konferenzpaper Nannen, Hauke; Zatocil, Heiko; Griepentrog, Gerd: Novel predictive start-up algorithm for soft starter driven induction motors, IECON 2020, 46th Annual Conference of the IEEE Industrial Electronics Society (IES), 18-21 October 2020, Singapore

[0002] Nannen et al

[0079] Vas, Peter: Sensorless Vector and Direct Torque Control. Monographs in Electrical and Electronic Engineering, Volume 42. Oxford: Oxford University Press 1998, ISBN-13: 978-0198564652

[0080]

Claims

[1] Method for operating a three-phase motor (4) on a multi-phase electrical network (5) by means of a motor control unit (1) with which one or more network phases (a, b, c) of the network (5) can each be switched by switching power electronic actuators (3), wherein in addition to a network-related switching criterion a rotor flux-related switching criterion is taken into account, taking into account a network-related shutdown criterion, where, for a switching decision to utilize the switching option of one or more power electronic actuators (3), both the grid-related switching criterion and the rotor flux-related switching criterion must be met, where either 2-phase or 3-phase switching is used, wherein in a space vector representation a 2-phase switching leads to a current vector that is fixed in its orientation and pulsating in its amplitude, and a 3-phase switching leads to a rotating current vector, wherein the network-related switching criterion is fulfilled if, in a space vector representation, the current vector of a switching possibility of two or more network phases (a, b, c) lies within a switching angle range (11) defined by a lower switching angle range limit (α) min ) and an upper limit of the switching angle range (α max ) is defined in relation to the mains voltage and / or mains current, and wherein the rotor flux-related switching criterion is fulfilled if, in a space vector representation, the current vector of a switching possibility of two or more network phases (a, b, c) lies within a flux and torque angle range (12) whose range limits (γ start ; γ end ) are defined in relation to the rotor flux. [2] Method according to claim 1, wherein the network-related disconnection criterion with respect to a power electronic actuator (3) that can switch a network phase (a, b, c) is satisfied when a current zero crossing occurs in this network phase (a, b, c). [3] Method according to one of the preceding claims, wherein the presence of an activation criterion is determined by means of a space vector representation of rotor flux (Ψ→2S), Mains voltage (U→1S) and stator current (I→1S) is being checked. [4] Method according to one of the preceding claims, wherein the flux and torque angle range (12) is by a rotation angle (δ rot ) is postponed to compensate for a time delay between a decision to switch on and the switchover itself. [5] Method according to any of the preceding claims, wherein the three-phase motor is an asynchronous motor, synchronous motor or PM line start motor. [6] Method according to one of the preceding claims, wherein the testing of grid-related and rotor flux-related switching criteria is preceded by a time period of flux build-up in which rotor flux is generated by switching on power electronic actuators. [7] Method according to one of the preceding claims, wherein only switching options are identified which result in negative torques and thus actively brake the three-phase motor (4). [8] Engine control unit (1) suitable for performing the steps of the method according to any one of claims 1 to 7. [9] Computer program product comprising commands that cause an engine control unit (1) according to claim 8 to perform the method steps according to any one of claims 1 to 7. [10] Computer-readable medium on which a computer program product according to claim 9 is stored.

Images