A position sensorless control method for a five-leg inverter dual permanent magnet motor system

Abstract

The present invention discloses a position sensorless control method for a five-arm inverter dual permanent magnet motor system. A five-arm inverter dual permanent magnet motor is set up, and the phase current is reconstructed by using an effective vector sampling method based on the DC bus current and the phase current of the two permanent magnet motors; a high-frequency voltage injection method is used to transfer the reference voltage vector in all cases to the non-phase current reconstruction blind area of the second or fifth sector, and then the effective vector sampling method is improved to reduce the phase current reconstruction error, and after reconstruction, the demodulation is extracted to achieve position sensorless control. The present invention installs a single current sensor on the DC bus to reconstruct the three-phase current, and then extracts the high-frequency component to estimate the rotor position to ensure the stable operation of the system. It not only realizes the phase current reconstruction but also estimates the permanent magnet motor rotor position. On the one hand, it improves the fault tolerance performance of the entire system, and on the other hand, it reduces the cost and volume of the system.

Description

Technical Field

[0001] The present invention relates to a control method for a permanent magnet motor in the field of coordinated control of multiple permanent magnet motors, and adopts a position sensorless control method for a five-bridge-arm inverter dual permanent magnet motor system. Background Art

[0002] With the continuous advancement of industrial technology, multi-PMM systems are widely used in fields such as electric vehicles and wind power generation. Consequently, the requirements for the control performance of multi-PMM systems are becoming increasingly stringent. Among multi-PMM systems, dual-PMM control systems are particularly widely used in industry and have attracted considerable research interest from scholars both domestically and internationally. Switching transistor failure is a major cause of inverter failure. In traditional dual-PMM systems, when a switching transistor in a single leg of a six-leg inverter fails, a five-leg inverter, as a superior fault-tolerant method, can independently control two PMMs. Using a five-leg inverter to drive two PMMs reduces the risk of system failure while also reducing system cost and size. This approach has been widely studied by numerous scholars in recent years.

[0003] In permanent magnet motor (PMM) control systems, accurate rotor position angle estimation is a key factor in ensuring stable system operation. Traditional PMM systems use position sensors to determine the motor's position angle, but this increases the cost and size of the entire system. Furthermore, these sensors are susceptible to external interference, which can affect the acquired position information. Therefore, to avoid these issues, the application of sensorless control technology to estimate the PMM rotor position is of great significance.

[0004] Accurate phase current information is another key factor in ensuring the stable operation of permanent magnet motor systems. Traditional permanent magnet motor systems require at least two current sensors to accurately measure the motor's three-phase currents. This not only increases system size and cost but can also cause measurement errors due to differences in current sensor parameters. Therefore, methods using a single current sensor to reconstruct the motor's phase currents have been widely studied. Summary of the Invention

[0005] To overcome the deficiencies of the prior art, the present invention proposes a position sensorless control method for a five-leg inverter dual permanent magnet motor system, which can achieve independent control of two three-phase permanent magnet motors while reducing the size and cost of the system, lowering the risk of system failure, and ensuring the stable operation of the permanent magnet motor system.

[0006] The specific scheme of the present invention is as follows:

[0007] In a five-leg inverter dual permanent magnet motor system, a duty cycle correction method can be used to output five-phase duty cycles to independently control the two permanent magnet motors. For the five-leg inverter dual permanent magnet motor, an effective vector sampling method is used to reconstruct the phase currents based on the corresponding relationship between the DC bus current in the five-leg inverter dual permanent magnet motor and the phase currents of the two permanent magnet motors under different switching states.

[0008] Due to non-ideal factors, all four sampling intervals must meet a minimum sampling time. Otherwise, if even one of them fails to meet the requirement, a phase current reconstruction blind zone will occur, making phase current reconstruction impossible. High-frequency voltage injection is used to shift the reference voltage vector in all cases to the non-phase current reconstruction blind zone in the second or fifth sector, avoiding the phase current reconstruction blind zone. Furthermore, by improving the reconstruction method of the effective vector sampling method, the phase current reconstruction error is reduced. After reconstruction, the reconstructed phase current is extracted and demodulated to achieve position sensorless control.

[0009] Under different switching states of the five-leg inverter, the current flow paths are different. The DC bus current contains different phase current information of the two permanent magnet motors. The five-leg inverter drives the dual permanent magnet motor system using a duty cycle correction method.

[0010] The effective vector sampling method is to set a current sensor on the DC bus of the five bridge arms in the five-bridge-arm inverter dual permanent magnet motor. Each control cycle includes four different switching states. The DC bus current is sampled by the current sensor during the action time of each switching state to obtain four DC bus currents; then, based on the relationship between each DC bus current and the phase currents of the two permanent magnet motors, two phase currents of each permanent magnet motor are obtained. Then, based on the fact that the sum of the three-phase currents of the permanent magnet motor stator windings is zero using star connection, the remaining third phase current of each permanent magnet motor is obtained, thereby realizing the reconstruction of the phase currents of the two permanent magnet motors.

[0011] The four different switching states are specifically states in which the upper bridge arm switch tubes of the bridge arms corresponding to different duty cycles in the five-phase duty cycle are turned on and the lower bridge arm switch tubes of the remaining bridge arms are turned on.

[0012] This method samples the DC bus current during the action time of four adjacent switching states within a control cycle, and then reconstructs the phase current based on the different phase current information of the two permanent magnet motors contained in the DC bus current under different switching states.

[0013] When the effective vector sampling method of the present invention is used to reconstruct the phase current, under ideal circumstances, the DC bus current sampling can be completed instantly and accurately to achieve phase current reconstruction.

[0014] However, in actual applications, due to the existence of non-ideal factors, DC bus current sampling cannot be completed instantaneously. The sampling requirements are not met in the sampling interval time of the DC bus current corresponding to the four switching states in each control cycle. DC bus current sampling cannot be completed accurately within a single control cycle, which will lead to unsuccessful phase current reconstruction and the occurrence of phase current reconstruction blind spots.

[0015] The method is to use high-frequency voltage injection to transfer the reference voltage vector in all cases to the non-phase current reconstruction blind area of the second or fifth sector to avoid the phase current reconstruction blind area, and use an improved effective vector sampling method to reconstruct the phase current. Then, the reconstructed phase current is used to extract the motor rotor position information to realize position sensorless control.

[0016] A permanent magnet motor has three phases: A, B, and C. If the angle of phase A is set to 0°, the angles of phase B and C are 120° and 240°, respectively. The area from 60° to 120° is the second sector, and the area from 240° to 300° is the fifth sector.

[0017] When the reference voltage vector is located in the second sector area, the action time of the two effective vectors used to synthesize the reference voltage vector is greater than 2T min The corresponding area is the non-phase current reconstruction blind area of the second sector.

[0018] When the reference voltage vector is located in the fifth sector, the action time of the two effective vectors used to synthesize the reference voltage vector is greater than 2T. min The corresponding area is the non-phase current reconstruction blind area of the fifth sector.

[0019] In specific implementation, the method uses the following formula to determine whether a phase current reconstruction blind zone occurs:

[0020] T min ≥T set +T on +T AD +T d

[0021] Where, T min Indicates the minimum time required for sampling, T set is the time it takes for the current to reach steady state due to the influence of the parasitic parameters of the switch tube and the inductance of the permanent magnet motor; T on is the on-time of the power switch tube; T AD is the delay time of the filter circuit and AD sampling; T d is the dead time.

[0022] The effective vector sampling method is specifically:

[0023] The following formula represents the relationship between the five bridge arm phase currents and the DC bus current obtained by four samplings at four different switching states. The five bridge arm phase currents are obtained by input processing:

[0024]

[0025] Among them, δ n is the duty cycle of the nth bridge arm, n represents the serial number of the bridge arm, n = 1, 2, 3, 4, 5, δ1~δ5 are the duty cycles of the five bridge arms arranged from large to small; i δn The duty cycle is δ n The phase current corresponding to the bridge arm; i sam1 、i sam2 、i sam3 、i sam4 They are the DC bus currents sampled during the action time of the four switching states.

[0026] The phase currents corresponding to the five-phase bridge arms are reconstructed using the above formula, and then the two-phase currents of each permanent magnet motor are reconstructed based on the phase currents corresponding to each bridge arm. The remaining last phase current is then obtained, thereby realizing the phase current reconstruction of the two permanent magnet motors.

[0027] Among them, i sam1 is the DC bus current sampled during the first switching state, i sam2 is the DC bus current sampled during the second switching state, i sam3 is the DC bus current sampled during the third switching state, i sam4 is the DC bus current sampled during the fourth switching state action time.

[0028] The first switching state is a state in which only the upper bridge arm switching tube of the bridge arm corresponding to the maximum duty cycle among the five-phase duty cycles is turned on, and the lower bridge arm switching tubes of the other four bridge arms are turned on; the second switching state is a state in which only the upper bridge arm switching tubes of the two bridge arms corresponding to the maximum and second largest duty cycles among the five-phase duty cycles are turned on, and the lower bridge arm switching tubes of the other three bridge arms are turned on; the third switching state is a state in which only the upper bridge arm switching tubes of the three bridge arms corresponding to the maximum, second largest and third largest duty cycles among the five-phase duty cycles are turned on, and the lower bridge arm switching tubes of the other two bridge arms are turned on; the fourth switching state is a state in which only the lower bridge arm switching tube of the bridge arm corresponding to the minimum duty cycle among the five-phase duty cycles is turned on, and the upper bridge arm switching tubes of the other four bridge arms are turned on.

[0029] The injected high-frequency voltage is injected into the three-phase lines of the two permanent magnet motors. The positive and negative values of the high-frequency voltage injected into each permanent magnet motor in each two consecutive control cycles are the same, that is, each permanent magnet motor alternately injects high-frequency voltages of different positive and negative values with two consecutive control cycles as a repeating unit, and the positive and negative values of the high-frequency voltages injected into the two permanent magnet motors in each control cycle are always opposite, and the amplitudes are always the same.

[0030] That is, the injection of high-frequency voltage is controlled according to the following formula:

[0031]

[0032] Among them, V h is the amplitude of the injected high-frequency voltage, T s Indicates the control cycle, K is 1, 2..., K represents the ordinal number of the control cycle, V inj Represents the injected high-frequency voltage signal.

[0033] The amplitude of the high-frequency voltage is obtained and set according to the following formula:

[0034]

[0035]

[0036] Among them, V h is the amplitude of the injected high-frequency voltage, V ref is the actual reference voltage of the permanent magnet motor, V dc is the DC bus voltage, V α and V β represents the stator voltage component of the α-axis and the stator voltage component of the β-axis, T min Indicates the minimum time required for sampling, T s Indicates the control period.

[0037] The magnitude of the high-frequency voltage injection amplitude is a key factor in ensuring the success of current reconstruction. The present invention sets the magnitude of the high-frequency voltage injection amplitude through formula calculation, which can effectively transfer all reference voltage vectors to the non-phase current reconstruction blind area of the second or fifth sector to achieve phase current reconstruction.

[0038] Through the above processing, different forms of high-frequency voltages are injected, and all reference voltage vectors are transferred to the non-phase current reconstruction blind area of the second or fifth sector. On the basis of ensuring the symmetry of the PWM wave, the four sampling intervals of the front and back half cycles meet the requirements, thereby realizing phase current reconstruction.

[0039] The improved effective vector sampling method is specifically as follows:

[0040] When the first permanent magnet motor is injected with a positive high-frequency voltage and the second permanent magnet motor is injected with a negative high-frequency voltage, the DC bus current i under the four switching states in the first half of the first control cycle of two consecutive control cycles is collected. sam11 、i sam12 、i sam13 、i sam14 When the first permanent magnet motor is injected with a positive high-frequency voltage and the second permanent magnet motor is injected with a negative high-frequency voltage, the DC bus current i under the four switching states in the second half of the first control cycle of two consecutive control cycles is collected. sam21 、i sam22 、i sam23 、i sam24 When the first permanent magnet motor is injected with a negative high-frequency voltage and the second permanent magnet motor is injected with a positive high-frequency voltage, the DC bus current i′ under the four switching states in the first half of the first control cycle of two consecutive control cycles is collected. sam11 , i′ sam12 , i′ sam13 , i′ sam14 When the first permanent magnet motor is injected with a negative high-frequency voltage and the second permanent magnet motor is injected with a positive high-frequency voltage, the DC bus current i′ under the four switching states in the second half of the first control cycle of two consecutive control cycles is collected. sam21 , i′ sam22 , i′ sam23 , i′ sam24 ;Then:

[0041] When a positive high-frequency voltage is injected into the first permanent magnet motor and a negative high-frequency voltage is injected into the second permanent magnet motor, the five bridge arm phase currents are obtained by inputting the DC bus currents according to the relationship between the five bridge arm phase currents and the DC bus currents obtained by multiple samplings in the same control cycle under four different switching states, as expressed by the following formula:

[0042]

[0043] When a negative high-frequency voltage is injected into the first permanent magnet motor and a positive high-frequency voltage is injected into the second permanent magnet motor, the five bridge arm phase currents are obtained by inputting the DC bus currents according to the relationship between the five bridge arm phase currents and the DC bus currents obtained by multiple samplings in the same control cycle with four different switching states, as expressed by the following formula:

[0044]

[0045] Among them, δ n is the duty cycle of the nth bridge arm, n represents the serial number of the bridge arm, n = 1, 2, 3, 4, 5, δ1~δ5 are the duty cycles of the five bridge arms arranged from large to small; iδn The duty cycle is δ n The phase current corresponding to the bridge arm; i sam11 、i sam12 、i sam13 、i sam14 are the DC bus currents under the first, second, third and fourth switching states when a positive high-frequency voltage is injected into the first permanent magnet motor, a negative high-frequency voltage is injected into the second permanent magnet motor, and the first half of the first control cycle in two consecutive control cycles; i sam21 、i sam22 、i sam23 、i sam24 are the DC bus currents under the first, second, third and fourth switching states when the first permanent magnet motor is injected with a positive high-frequency voltage, the second permanent magnet motor is injected with a negative high-frequency voltage, and the second half of the first control cycle in two consecutive control cycles; i′ sam11 , i′ sam12 , i′ sam13 , i′ sam14 are the DC bus currents under the first, second, third and fourth switching states when the first permanent magnet motor is injected with a negative high-frequency voltage, the second permanent magnet motor is injected with a positive high-frequency voltage, and the first half of the first control cycle in two consecutive control cycles; i′ sam21 , i′ sam22 , i′ sam23 , i′ sam24 They are respectively the DC bus currents under the first, second, third and fourth switching states when a negative high-frequency voltage is injected into the first permanent magnet motor, a positive high-frequency voltage is injected into the second permanent magnet motor, and the second half of the first control cycle in two consecutive control cycles.

[0046] The first switching state is a state in which only the upper bridge arm switching tube of the bridge arm corresponding to the maximum duty cycle among the five-phase duty cycles is turned on, and the lower bridge arm switching tubes of the other four bridge arms are turned on; the second switching state is a state in which only the upper bridge arm switching tubes of the two bridge arms corresponding to the maximum and second largest duty cycles among the five-phase duty cycles are turned on, and the lower bridge arm switching tubes of the other three bridge arms are turned on; the third switching state is a state in which only the upper bridge arm switching tubes of the three bridge arms corresponding to the maximum, second largest and third largest duty cycles among the five-phase duty cycles are turned on, and the lower bridge arm switching tubes of the other two bridge arms are turned on; the fourth switching state is a state in which only the lower bridge arm switching tube of the bridge arm corresponding to the minimum duty cycle among the five-phase duty cycles is turned on, and the upper bridge arm switching tubes of the other four bridge arms are turned on.

[0047] While high-frequency injection can address the blind spot problem of reconstructing phase currents, it can also lead to large current fluctuations within a cycle. Furthermore, because the actual current sampling points differ from the reconstructed current sampling points, the reconstructed current errors are large. The present invention improves upon the high-frequency injection method described above by adding a high-frequency component to the reconstructed current and altering the injection period. This significantly increases the high-frequency component of the reconstructed current, thereby improving the accuracy of the rotor position estimate.

[0048] The present invention extracts the high-frequency component of the reconstructed phase current by combining high-frequency voltage injection, then demodulates it, and finally estimates the rotor position of the permanent magnet motor through an orthogonal phase-locked loop. The rotor position of the permanent magnet motor is used for control, specifically:

[0049] First, establish the initial voltage model of the permanent magnet motor in a two-phase stationary coordinate system:

[0050]

[0051] L1=(L dh +L qh ) / 2, L2=(L dh -L qh ) / 2

[0052] Where L1 is the mean inductance of the permanent magnet motor, L2 is the differential inductance of the permanent magnet motor, v αh 、v βh are the α-axis and β-axis high-frequency voltage components of the permanent magnet motor in the two-phase stationary coordinate system, i αh 、i βh are the α-axis and β-axis high-frequency response current components of the permanent magnet motor in the two-phase stationary coordinate system; p represents the differential factor, L dh , L qh are the high-frequency inductance values of the d-axis and q-axis, θ e Indicates the rotor position of the permanent magnet motor;

[0053] According to the injected high-frequency voltage signal, the injected voltage model of the permanent magnet motor after the high-frequency voltage is injected in the two-phase stationary coordinate system is established:

[0054]

[0055]

[0056] Where i represents the serial number of the two permanent magnet motors, i=1, 2; v αhi 、v βhi They represent the high-frequency voltage components of the α-axis and β-axis of the i-th permanent magnet motor in the two-phase stationary coordinate system;

[0057] According to the initial voltage model and the injected voltage model, the high-frequency response current change of the permanent magnet motor is obtained after reprocessing:

[0058]

[0059]

[0060] Δt=2T s (14)

[0061] Among them, i αhi and i βhi They represent the high-frequency response current components of the α-axis and β-axis of the i-th permanent magnet motor in the two-phase stationary coordinate system, Δi αhi and Δi βhi are the high-frequency response current changes of the α-axis and β-axis of the permanent magnet motor in the two-phase stationary coordinate system, V inj represents the injected high-frequency voltage signal, θ ei represents the rotor position of the i-th permanent magnet motor, Δt represents the time change, which is two control cycles;

[0062] According to the above formula of the high-frequency response current change of the permanent magnet motor, signal processing is performed and multiplied by the sign function to obtain the following formula to calculate the rotor position θ of the permanent magnet motor: ei , realize position sensorless control and realize system closed-loop control:

[0063]

[0064]

[0065]

[0066] Among them, I sini represents the high-frequency response current envelope of the α-axis of the i-th permanent magnet motor, I cosi Represents the high-frequency response current envelope of the β-axis of the i-th permanent magnet motor.

[0067] The five-bridge-arm inverter dual permanent magnet motor includes two permanent magnet motors and an inverter consisting of five bridge arms. The five bridge arms are connected in parallel, wherein the midpoints of three bridge arms are connected to the three-phase line of one of the permanent magnet motors, and the midpoints of three other bridge arms are connected to the three-phase line of the other permanent magnet motor. Among the three bridge arms connected to the two permanent magnet motors, only one bridge arm is shared as a common bridge arm.

[0068] Each bridge arm is composed of two switching tubes connected in series.

[0069] The characteristics and beneficial effects of the present invention are:

[0070] The present invention is based on a five-leg inverter dual permanent magnet motor system. By installing a single current sensor on the DC bus, the three-phase currents of the two permanent magnet motors are reconstructed. The reconstructed phase currents are then processed to extract high-frequency components and estimate the rotor positions of the two permanent magnet motors to ensure stable operation of the system.

[0071] The present invention realizes both phase current reconstruction and estimation of the permanent magnet motor rotor position in only one way, which improves the fault tolerance performance of the entire system on the one hand and reduces the cost and volume of the system on the other hand. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 This is the topological structure diagram of the five-leg inverter and two permanent magnet motors;

[0073] Figure 2 This is a diagram of the high-frequency voltage signal injection method;

[0074] Figure 3 This is the principle diagram of high-frequency voltage injection;

[0075] Figure 4 It is the principle block diagram of the orthogonal phase-locked loop;

[0076] Figure 5 This is the simulation result diagram of the first permanent magnet motor running steadily at 20 r / min with a load of 3 Nm applied to both permanent magnet motors;

[0077] Figure 6 The first permanent magnet motor simulation result diagram shows that the speed of the permanent magnet motor accelerates from 20r / min to 50r / min from startup to 1s, and decelerates from 50r / min to 20r / min from 1s to 2s. The second permanent magnet motor runs stably at 20r / min. A load of 3Nm is applied to both permanent magnet motors.

[0078] Figure 7 The first permanent magnet motor simulation results are as follows: the first permanent magnet motor is loaded with 3Nm and starts at 20r / min. After 1.2s, the load is stepped to 5Nm. After 2.4s, the load is reduced to 3Nm. The second permanent magnet motor is loaded with 3Nm and runs steadily at 20r / min.

[0079] Figure 8 The following figure shows the simulation results of two permanent magnet motors running stably at 20 r / min with a load of 3 Nm applied to both motors.

[0080] Figure 9The simulation results of the two permanent magnet motors are as follows: the speed of the first permanent magnet motor accelerates from 20r / min to 50r / min from startup to 1s, and decelerates from 50r / min to 20r / min from 1s to 2s. The second permanent magnet motor runs stably at 20r / min. A load of 3Nm is applied to both permanent magnet motors.

[0081] Figure 10 This is the simulation result diagram of two permanent magnet motors: the first permanent magnet motor is loaded with 3Nm and starts at 20r / min. After 1.2s, the load is stepped to 5Nm, and after 2.4s, the load is reduced to 3Nm. The second permanent magnet motor is loaded with 3Nm and runs stably at 20r / min. DETAILED DESCRIPTION

[0082] The present invention will be further described below with reference to the accompanying drawings and specific implementations.

[0083] This invention reconstructs the three-phase currents of two permanent magnet motors by installing a single current sensor on the DC bus. The reconstructed currents are then processed to extract high-frequency components to estimate the rotor positions of the two permanent magnet motors. The reconstructed currents and estimated positions are then used for closed-loop control. This invention employs a duty cycle correction method for a five-leg inverter dual permanent magnet motor system to achieve phase current reconstruction and sensorless control. The invention also details methods for resolving phase current blind spots, phase current reconstruction methods, calculation of high-frequency voltage injection amplitudes, and sensorless control methods.

[0084] The specific implementation of the present invention is as follows:

[0085] 1. Five-leg inverter dual permanent magnet motor system and its phase current reconstruction technology

[0086] The five-bridge-arm inverter dual permanent magnet motor includes two permanent magnet motors and an inverter consisting of five bridge arms. The five bridge arms are connected in parallel, with the midpoints of three bridge arms connected to the three-phase line of one permanent magnet motor, and the midpoints of three bridge arms connected to the three-phase line of the other permanent magnet motor. There is only one bridge arm shared by the three bridge arms connected to the two permanent magnet motors as a common bridge arm. Each bridge arm is composed of two switching tubes connected in series, and its topology is as follows: Figure 1 In the figure, A, B, and C are the three bridge arms of inverter 1, and E, D, and C are the three bridge arms of inverter 2. The duty cycle correction method derives a formula to equalize the duty cycles of the common phases. The output five-phase duty cycle independently controls the two permanent magnet motors, achieving true decoupling and improving DC bus voltage utilization.

[0087] like Figure 1As shown, a single current sensor is installed on the DC bus to reconstruct the three-phase currents of two permanent magnet motors. The relationship between each DC bus current and the two permanent magnet motor phase currents is used to determine two of the motor's phase currents. Furthermore, the remaining third phase current of each motor is determined by the fact that the sum of the three-phase currents in the star-connected stator windings is zero, thus reconstructing the phase currents of the two permanent magnet motors.

[0088]

[0089] Among them, δ n is the duty cycle of the nth bridge arm, n represents the serial number of the bridge arm, n = 1, 2, 3, 4, 5, δ1~δ5 are the duty cycles of the five bridge arms arranged from large to small; i δn The duty cycle is δ n The phase current corresponding to the bridge arm; i sam1 、i sam2 、i sam3 、i sam4 They are the DC bus currents sampled during the action time of the four switching states.

[0090] However, due to non-ideal factors in actual applications, each sampling interval must meet a minimum sampling time T min :

[0091] T min ≥T set +T on +T AD +T d (2)

[0092] Where, T min Indicates the minimum time required for sampling, T set is the time it takes for the current to reach steady state due to the influence of the parasitic parameters of the switch tube and the inductance of the permanent magnet motor; T on is the on-time of the power switch tube; T AD is the delay time of the filter circuit and AD sampling; T d is the dead time.

[0093] In the five-leg inverter dual permanent magnet motor system, the three-phase currents of the two permanent magnet motors can only be reconstructed when all four sampling intervals meet the requirements. Unlike the single permanent magnet motor system, the phase current reconstruction of the five-leg inverter dual permanent magnet motor system will also produce new phase current reconstruction blind spots.

[0094] 2. Position sensorless control of a five-leg inverter dual permanent magnet motor system based on a single current sensor

[0095] By adopting the high-frequency voltage injection method, the reference voltage vector in all cases is transferred to the non-phase current reconstruction blind area of the second or fifth sector. Then, the reconstruction error is reduced by improving the effective vector sampling method. Finally, the reconstructed phase current is processed to extract the high-frequency component to realize position sensorless control.

[0096] 2.1 High-frequency voltage injection method

[0097] The present invention combines the characteristics of the five-leg inverter to select different forms of high-frequency voltages to be injected into the three-phase lines of the two permanent magnet motors to avoid the blind area of the phase current reconstruction and realize the phase current reconstruction. The injection form is as follows: Figure 2 When a positive voltage is injected into the three-phase line of the first permanent magnet motor, the reference voltage vector can be transferred to the second sector non-phase current reconstruction blind area, and the three-phase duty cycle is δ b1 >δ a1 >δ c1 .

[0098] Injecting a negative voltage into the three-phase line of the second permanent magnet motor can pull the reference voltage vector to the non-phase current reconstruction blind area of the fifth sector, and the three-phase duty cycle is δ c2 >δ a2 >δ b2 According to the duty cycle correction method, the duty cycles of the common phase c are equal, so the duty cycle relationship of the six phases of the two permanent magnet motors is δ b1 >δ a1 >δ c1 =δ c2 >δ a2 >δ b2 On the contrary, when negative voltage is injected into the three-phase line of the first permanent magnet motor and positive voltage is injected into the three-phase line of the second permanent magnet motor, the analysis is similar to the above, and the phase current reconstruction blind area problem is finally solved by this method.

[0099] 2.2 Phase current reconstruction method

[0100] To address the problem that high-frequency voltage injection causes large current fluctuations within a cycle and large reconstructed current errors, the present invention changes the reconstruction method.

[0101] In one control cycle, the method of taking an average of each phase current in the first and second half cycles is adopted. This method can reduce the phase current reconstruction error to a certain extent. In order to improve the high-frequency component of the reconstructed current, the injection method is changed. Each permanent magnet motor alternately injects high-frequency voltages of different positive and negative values in two consecutive control cycles as a repetitive unit. In each control cycle, the positive and negative values of the high-frequency voltages injected by the two permanent magnet motors are always opposite, and the amplitudes are always the same. This greatly increases the high-frequency component of the reconstructed phase current, which can better estimate the rotor position of the permanent magnet motor. Its injection method and sampling method are as follows: Figure 2 shown.

[0102] When the first permanent magnet motor is injected with a positive high-frequency voltage and the second permanent magnet motor is injected with a negative high-frequency voltage, the DC bus current i under the four switching states in the first half of the first control cycle of two consecutive control cycles is collected. sam11 、i sam12 、i sam13 、i sam14 When the first permanent magnet motor is injected with a positive high-frequency voltage and the second permanent magnet motor is injected with a negative high-frequency voltage, the DC bus current i under the four switching states in the second half of the first control cycle of two consecutive control cycles is collected. sam21 、i sam22 、i sam23 、i sam24 When the first permanent magnet motor is injected with a negative high-frequency voltage and the second permanent magnet motor is injected with a positive high-frequency voltage, the DC bus current i′ under the four switching states in the first half of the first control cycle of two consecutive control cycles is collected. sam11 , i′ sam12 , i′ sam13 , i′ sam14 When the first permanent magnet motor is injected with a negative high-frequency voltage and the second permanent magnet motor is injected with a positive high-frequency voltage, the DC bus current i′ under the four switching states in the second half of the first control cycle of two consecutive control cycles is collected. sam21 , i′ sam22 , i′ sam23 , i′ sam24 ;

[0103] Then, when a positive high-frequency voltage is injected into the first permanent magnet motor and a negative high-frequency voltage is injected into the second permanent magnet motor, the five bridge arm phase currents are obtained by inputting the DC bus currents according to the relationship between the five bridge arm phase currents and the DC bus currents obtained by multiple samplings of the four different switching states in the same control cycle, as expressed by the following formula:

[0104]

[0105] When a negative high-frequency voltage is injected into the first permanent magnet motor and a positive high-frequency voltage is injected into the second permanent magnet motor, the relationship between the five bridge arm phase currents and the DC bus currents obtained by multiple samplings in the same control cycle at four different switching states is expressed by the following formula. The input DC bus currents are processed to obtain the phase currents of the five bridge arms:

[0106]

[0107] Taking the b1 phase current as an example, a simple analysis is made on why this method can reduce the reconstructed current error.b1 Fluctuation curve Figure 2 As shown in the figure. b1_con 、i b1_re1 and i b1_re2 They are respectively the b1 phase current sampled by the actual current sensor when the first permanent magnet motor is injected with positive voltage, the b1 phase current reconstructed in the first half cycle, and the b1 phase current reconstructed in the second half cycle; i′ b1_con , i′ b1_re1 and i′ b1_re2 They are respectively the b1 phase current sampled by the actual current sensor when the negative voltage is injected into the first permanent magnet motor, the b1 phase current reconstructed in the first half cycle, and the b1 phase current reconstructed in the second half cycle; the fundamental frequency currents i1, i2, and i3 obtained after actual high frequency elimination, high frequency elimination by sampling in the first half cycle, and high frequency elimination by sampling in the second half cycle are:

[0108]

[0109] The error Δ in reconstructing the current in the first and second half cycles e1 , Δ e2 for:

[0110]

[0111] The current fluctuation is mainly due to the influence of high-frequency injection, and it is assumed that the fundamental frequency current remains basically unchanged.

[0112] Due to different sampling times, sample i′ from the first half cycle b1_re1 Than i b1_re1 The change is large, sample i from the second half of the cycle b1_re2 Than i′ b1_re2 The change is large, so Δ e1 >0, Δ e2 <0, the error of the reconstructed current by the improved method is 0.5(Δ e1 +Δ e2 ), so in principle it can be seen that the reconstruction error can be effectively reduced.

[0113] 2.3 High-frequency voltage injection amplitude

[0114] The magnitude of the high-frequency voltage injection amplitude is the key factor to ensure the success of current reconstruction. The injection principle is as follows: Figure 3 shown.

[0115] The amplitude of the high-frequency voltage is obtained and set according to the following formula:

[0116]

[0117]

[0118] Among them, V h is the amplitude of the injected high-frequency voltage, Vref is the actual reference voltage of the permanent magnet motor, V dc is the DC bus voltage, V α and V β represents the stator voltage component of the α-axis and the stator voltage component of the β-axis, T min Indicates the minimum time required for sampling, T s Indicates the control period.

[0119] 2.4. Position sensorless control

[0120] The specific implementation combines the proposed injection method to extract the high-frequency component of the reconstructed current and realize position estimation.

[0121] Establish the voltage model of permanent magnet synchronous permanent magnet motor:

[0122]

[0123] Where v d 、v q is the stator voltage component of the dq axis; R s is the stator resistance; L d , L q is the dq axis stator inductance; i d 、i q is the stator current component of the dq axis; ω e is the electrical angular velocity of the permanent magnet motor; ψ f is the permanent magnet flux.

[0124] When the permanent magnet motor runs at zero or low speed, after high-frequency voltage is injected into the permanent magnet motor, the stator resistance voltage, rotation voltage and back electromotive force can be ignored, and the voltage model can be simplified as follows:

[0125]

[0126] Where, v dh 、v qh 、i dh 、i qh , L dh , L qh They are the high-frequency voltage signal, high-frequency current response component and high-frequency inductance value injected into the d-axis and q-axis respectively.

[0127] Convert the above voltage model to the stationary coordinate system and establish the initial voltage model of the permanent magnet motor in the two-phase stationary coordinate system:

[0128]

[0129] L1=(L dh +L qh ) / 2, L2=(L dh -Lqh ) / 2 (12)

[0130] Where L1 is the mean inductance of the permanent magnet motor, L2 is the differential inductance of the permanent magnet motor, v αh 、v βh The high-frequency square wave voltage components of the α-axis and β-axis, i αh 、i βh The α-axis and β-axis high-frequency response current components respectively; p represents the differential factor; L dh , L qh are the high-frequency inductance values of the d-axis and q-axis respectively; θ e Indicates the rotor position of the permanent magnet motor.

[0131] According to the injected high-frequency voltage signal, the injected voltage model of the permanent magnet motor after the high-frequency voltage is injected in the two-phase stationary coordinate system is established:

[0132]

[0133]

[0134] Where i represents the serial number of the two permanent magnet motors, i=1, 2; v αhi 、v βhi Respectively represent the high-frequency voltage components of the α-axis and β-axis of the i-th permanent magnet motor; V h is the amplitude of the injected high-frequency voltage, T s Indicates the control cycle, K is 1, 2..., K represents the ordinal number of the control cycle, V inj Represents the injected high-frequency voltage signal.

[0135] According to the initial voltage model and the injected voltage model, the high-frequency response current change of the permanent magnet motor is obtained after reprocessing:

[0136]

[0137]

[0138] Δt=2T s (17)

[0139] Among them, i αhi and i βhi They represent the high-frequency response currents of the α-axis and β-axis of the i-th motor, Δi αhi and Δi βhi Respectively represent the high-frequency response current changes of the α-axis and β-axis of the i-th motor, V inj represents the injected high-frequency square wave voltage signal, θ ei It represents the rotor position of the i-th permanent magnet motor, Δt represents the time change, which is two control cycles.

[0140] According to the above formula of the high-frequency response current change of the permanent magnet motor, the signal is processed and multiplied by the sign function to obtain the following formula, which is used to calculate the rotor positions of the two permanent magnet motors, realize position sensorless control, and realize system closed-loop control:

[0141]

[0142]

[0143]

[0144] Among them, I sini represents the high-frequency response current envelope of the α-axis of the i-th permanent magnet motor, I cosi Represents the high-frequency response current envelope of the β-axis of the i-th permanent magnet motor.

[0145] Finally, the above formula is passed Figure 4 The orthogonal phase-locked loop shown estimates the permanent magnet motor rotor position and performs system closed-loop control.

[0146] To verify the correctness and effectiveness of the proposed method, a simulation was performed. In the simulation diagram, the first permanent magnet motor is represented by PMSM1, and the second permanent magnet motor is represented by PMSM2. The two permanent magnet motors are of the same model, and their parameters are shown in Table 1.

[0147] Table 1 Permanent magnet motor parameters

[0148]

[0149]

[0150] (1) Single permanent magnet motor position sensorless control

[0151] In the simulation, the first permanent magnet motor is closed-loop controlled by reconstructed three-phase current and estimated permanent magnet motor rotor position, and the second permanent magnet motor is closed-loop controlled by actual three-phase current and permanent magnet motor rotor position.

[0152] (1) Steady-state performance

[0153] The specific working condition is that both permanent magnet motors run at 20r / min under 3Nm load. Figure 5 For its simulation results. Figure 5 (a) From top to bottom, the actual b-phase current, the reconstructed b-phase current, and the b-phase current reconstruction error of the first permanent magnet motor. Figure 5(b) From top to bottom, the actual position, estimated position, and position estimation error of the first permanent magnet motor are shown. The figure shows that the reconstructed phase current of the first permanent magnet motor is consistent with the actual phase current, and the actual rotor position of the permanent magnet motor is consistent with the estimated rotor position of the permanent magnet motor. The reconstructed three-phase current and estimated rotor position of the permanent magnet motor are used for closed-loop control of the first permanent magnet motor, ensuring stable operation of the permanent magnet motor.

[0154] (2) Dynamic performance

[0155] The first working condition is that the speed of the first permanent magnet motor accelerates from 20r / min to 50r / min from the start to 1s, and decelerates from 50r / min to 20r / min from 1s to 2s. The second permanent magnet motor runs stably at 20r / min. A load of 3Nm is applied to both permanent magnet motors. Figure 6 For its simulation results. Figure 6 (a) From top to bottom, the actual b-phase current, the reconstructed b-phase current, and the b-phase current reconstruction error of the first permanent magnet motor. Figure 6 (b) From top to bottom, the actual and estimated positions of the first permanent magnet motor, along with the position estimation error, are shown. The figure shows that the reconstructed phase currents of the first permanent magnet motor are consistent with the actual phase currents, and the actual and estimated rotor positions of the permanent magnet motor are consistent. The reconstructed three-phase currents and estimated rotor positions are used for closed-loop control of the first permanent magnet motor, ensuring stable operation.

[0156] The second working condition is that the first permanent magnet motor is loaded with 3Nm and starts at 20r / min. After 1.2s, the load is stepped to 5Nm. After 2.4s, the load is reduced to 3Nm. The second permanent magnet motor is loaded with 3Nm and runs stably at 20r / min. Figure 7 For its simulation results. Figure 7 (a) From top to bottom, the actual b-phase current, the reconstructed b-phase current, and the b-phase current reconstruction error of the first permanent magnet motor. Figure 7 (b) From top to bottom, the actual and estimated positions of the first permanent magnet motor, along with the position estimation error, are shown. The figure shows that the reconstructed phase currents of the first permanent magnet motor are consistent with the actual phase currents, and the actual and estimated rotor positions of the permanent magnet motor are consistent. The reconstructed three-phase currents and estimated rotor positions are used for closed-loop control of the first permanent magnet motor, ensuring stable operation.

[0157] (2) Dual permanent magnet motor position sensorless control

[0158] In the simulation, both permanent magnet motors are controlled in a closed loop through the reconstructed three-phase current and the estimated permanent magnet motor rotor position.

[0159] (1) Steady-state performance

[0160] The specific working condition is that both permanent magnet motors run at 20r / min under 3Nm load. Figure 8 For its simulation results. Figure 8 (a) From top to bottom, the actual b-phase current, the reconstructed b-phase current, and the b-phase current reconstruction error of the two permanent magnet motors. Figure 8 (b) From top to bottom, the actual and estimated positions of the two permanent magnet motors, along with the position estimation errors, are shown. The figure shows that the reconstructed phase currents of the dual permanent magnet motors are consistent with the actual phase currents, and the actual and estimated permanent magnet motor rotor positions are consistent. The reconstructed three-phase currents and estimated permanent magnet motor rotor positions are used for closed-loop control of the two permanent magnet motors, ensuring stable operation.

[0161] (2) Dynamic performance

[0162] The first working condition is that the speed of the first permanent magnet motor accelerates from 20r / min to 50r / min from the start to 1s, and decelerates from 50r / min to 20r / min from 1s to 2s. The second permanent magnet motor runs stably at 20r / min. A load of 3Nm is applied to both permanent magnet motors. Figure 9 For its simulation results. Figure 9 (a) From top to bottom, the actual b-phase current, the reconstructed b-phase current, and the b-phase current reconstruction error of the two permanent magnet motors. Figure 9 (b) From top to bottom, the actual and estimated positions of the two permanent magnet motors, along with the position estimation errors, are shown. The figure shows that the reconstructed phase currents of the two permanent magnet motors are consistent with the actual phase currents, and the actual and estimated rotor positions of the permanent magnet motors are consistent. The reconstructed three-phase currents and estimated rotor positions are used for closed-loop control of the two permanent magnet motors, ensuring stable operation.

[0163] The second working condition is that the first permanent magnet motor is loaded with 3Nm and starts at 20r / min. After 1.5s, the load is stepped to 5Nm. After 2.2s, the load is reduced to 3Nm. The second permanent magnet motor is loaded with 3Nm and runs stably at 20r / min. Figure 10 For its simulation results. Figure 10 (a) From top to bottom, the actual b-phase current, the reconstructed b-phase current, and the b-phase current reconstruction error of the two permanent magnet motors. Figure 10(b) From top to bottom, the actual and estimated positions of the two permanent magnet motors, along with the position estimation errors, are shown. The figure shows that the reconstructed phase currents of the two permanent magnet motors are consistent with the actual phase currents, and the actual and estimated rotor positions of the permanent magnet motors are consistent. The reconstructed three-phase currents and estimated rotor positions are used for closed-loop control of the dual permanent magnet motors, ensuring stable operation of the motors.

[0164] Ultimately, the present invention injects a high-frequency voltage to shift the reference voltage vector to the non-blind zone of the second or fifth sector, avoiding the phase current reconstruction blind zone and enabling phase current reconstruction. The reconstructed phase current is then processed to extract the high-frequency component, enabling sensorless control. This single approach achieves both phase current reconstruction and permanent magnet motor rotor position estimation, improving the fault tolerance and reliability of the entire system.

Claims

1. A position sensorless control method for a five-leg inverter dual permanent magnet motor system, characterized by: For the five-leg inverter dual permanent magnet motor, the phase current is reconstructed using the effective vector sampling method based on the corresponding relationship between the DC bus current in the five-leg inverter dual permanent magnet motor and the phase currents of the two permanent magnet motors under different switching states. High-frequency voltage injection is used to transfer the reference voltage vector in all cases to the non-phase current reconstruction blind zone of the second or fifth sector to avoid the phase current reconstruction blind zone. Then, the effective vector sampling method is improved to reduce the phase current reconstruction error. After reconstruction, the reconstructed phase current is extracted and demodulated to achieve position sensorless control. The effective vector sampling method is to set a current sensor on the DC bus of the five bridge arms in the five-bridge-arm inverter dual permanent magnet motor. Each control cycle includes four different switching states. The DC bus current is sampled by the current sensor during the action time of each switching state to obtain four DC bus currents; then, based on the relationship between each DC bus current and the phase currents of the two permanent magnet motors, two phase currents of each permanent magnet motor are obtained. Then, based on the fact that the sum of the three-phase currents of the permanent magnet motor stator windings is zero using star connection, the remaining third phase current of each permanent magnet motor is obtained, thereby realizing the reconstruction of the phase currents of the two permanent magnet motors.

2. The method for controlling a five-leg inverter dual permanent magnet motor system without position sensors according to claim 1, characterized in that: The method is to use high-frequency voltage injection to transfer the reference voltage vector in all cases to the non-phase current reconstruction blind area of the second or fifth sector to avoid the phase current reconstruction blind area, and use an improved effective vector sampling method to reconstruct the phase current. Then, the reconstructed phase current is used to extract the motor rotor position information to realize position sensorless control.

3. A position sensorless control method for a five-leg inverter dual permanent magnet motor system according to claim 1 or 2, characterized in that: The effective vector sampling method is specifically: The following formula represents the relationship between the five bridge arm phase currents and the DC bus current obtained by four samplings at four different switching states. The five bridge arm phase currents are obtained by input processing: Among them, δ n is the duty cycle of the nth bridge arm, n represents the serial number of the bridge arm, n = 1, 2, 3, 4, 5, δ1~δ5 are the duty cycles of the five bridge arms arranged from large to small; i δn The duty cycle is δ n The phase current corresponding to the bridge arm; i sam1 、i sam2 、i sam3 、i sam4 They are the DC bus currents sampled during the action time of the four switching states.

4. The position sensorless control method for a five-leg inverter dual permanent magnet motor system according to claim 3, characterized in that: The high-frequency voltage injection method is to inject the high-frequency voltage into the three-phase lines of the two permanent magnet motors, and the positive and negative values of the high-frequency voltage injected into each permanent magnet motor in every two consecutive control cycles are the same.

5. A position sensorless control method for a five-leg inverter dual permanent magnet motor system according to claim 3 or 4, characterized in that: The amplitude of the high-frequency voltage is obtained and set according to the following formula: Among them, V h is the amplitude of the injected high-frequency voltage, V ref is the actual reference voltage of the permanent magnet motor, V dc is the DC bus voltage, V α and V β represents the stator voltage component of the α-axis and the stator voltage component of the β-axis, T min Indicates the minimum time required for sampling, T s Indicates the control period.

6. A position sensorless control method for a five-leg inverter dual permanent magnet motor system according to claim 3 or 4, characterized in that: The improved effective vector sampling method is specifically as follows: When the first permanent magnet motor is injected with a positive high-frequency voltage and the second permanent magnet motor is injected with a negative high-frequency voltage, the DC bus current i under the four switching states in the first half of the first control cycle of two consecutive control cycles is collected. sam11 、i sam12 、i sam13 、i sam14 When the first permanent magnet motor is injected with a positive high-frequency voltage and the second permanent magnet motor is injected with a negative high-frequency voltage, the DC bus current i under the four switching states in the second half of the first control cycle of two consecutive control cycles is collected. sam21 、i sam22 、i sam23 、i sam24 When the first permanent magnet motor is injected with a negative high-frequency voltage and the second permanent magnet motor is injected with a positive high-frequency voltage, the DC bus current i′ under the four switching states in the first half of the first control cycle of two consecutive control cycles is collected. sam11 , i′ sam12 , i′ sam13 , i′ sam14 When the first permanent magnet motor is injected with a negative high-frequency voltage and the second permanent magnet motor is injected with a positive high-frequency voltage, the DC bus current i′ under the four switching states in the second half of the first control cycle of two consecutive control cycles is collected. sam21 , i′ sam22 , i′ sam23 , i′ sam24 ;Then: When a positive high-frequency voltage is injected into the first permanent magnet motor and a negative high-frequency voltage is injected into the second permanent magnet motor, the five bridge arm phase currents are obtained by input processing based on the relationship between the five bridge arm phase currents and the DC bus currents obtained by multiple samplings in the same control cycle with four different switching states, as expressed by the following formula: When a negative high-frequency voltage is injected into the first permanent magnet motor and a positive high-frequency voltage is injected into the second permanent magnet motor, the five bridge arm phase currents are obtained by input processing based on the relationship between the five bridge arm phase currents and the DC bus currents obtained by multiple samplings in the same control cycle with four different switching states, as expressed by the following formula: Among them, δ n is the duty cycle of the nth bridge arm, n represents the serial number of the bridge arm, n = 1, 2, 3, 4, 5, δ1~δ5 are the duty cycles of the five bridge arms arranged from large to small; i δn The duty cycle is δ n The phase current corresponding to the bridge arm; i sam11 、i sam12 、i sam13 、i sam14 are the DC bus currents under the first, second, third and fourth switching states when a positive high-frequency voltage is injected into the first permanent magnet motor, a negative high-frequency voltage is injected into the second permanent magnet motor, and the first half of the first control cycle in two consecutive control cycles; i sam21 、i sam22 、i sam23 、i sam24 are the DC bus currents under the first, second, third and fourth switching states when the first permanent magnet motor is injected with a positive high-frequency voltage, the second permanent magnet motor is injected with a negative high-frequency voltage, and the second half of the first control cycle in two consecutive control cycles; i′ sam11 , i′ sam12 , i′ sam13 , i′ sam14 are the DC bus currents under the first, second, third and fourth switching states when the first permanent magnet motor is injected with a negative high-frequency voltage, the second permanent magnet motor is injected with a positive high-frequency voltage, and the first half of the first control cycle in two consecutive control cycles; i′ sam21 , i′ sam22 , i′ sam23 , i′ sam24 They are respectively the DC bus currents under the first, second, third and fourth switching states when a negative high-frequency voltage is injected into the first permanent magnet motor, a positive high-frequency voltage is injected into the second permanent magnet motor, and the second half of the first control cycle in two consecutive control cycles.

7. A position sensorless control method for a five-leg inverter dual permanent magnet motor system according to claim 3 or 4, characterized in that: Combined with the injection of high-frequency voltage, the high-frequency component of the reconstructed phase current is extracted and then demodulated. Finally, the rotor position of the permanent magnet motor is estimated through an orthogonal phase-locked loop, and the rotor position of the permanent magnet motor is used for control. Specifically: First, establish the initial voltage model of the permanent magnet motor in a two-phase stationary coordinate system: L1=(L dh +L qh ) / 2,L2=(L dh -L qh ) / 2 Where L1 is the mean inductance of the permanent magnet motor, L2 is the differential inductance of the permanent magnet motor, and v αh 、v βh are the α-axis and β-axis high-frequency voltage components of the permanent magnet motor in the two-phase stationary coordinate system, i αh 、i βh are the α-axis and β-axis high-frequency response current components of the permanent magnet motor in the two-phase stationary coordinate system; p represents the differential factor, L dh , L qh are the high-frequency inductance values of the d-axis and q-axis, θ e Indicates the rotor position of the permanent magnet motor; Establish the injected voltage model of the permanent magnet motor after injecting high-frequency voltage in the two-phase stationary coordinate system: Where i represents the serial number of the two permanent magnet motors, i=1, 2; v αhi 、v βhi They represent the high-frequency voltage components of the α-axis and β-axis of the i-th permanent magnet motor in the two-phase stationary coordinate system; According to the initial voltage model and the injected voltage model, the high-frequency response current change of the permanent magnet motor is obtained after reprocessing: Δt=2T s (14) Among them, i αhi and i βhi They represent the high-frequency response current components of the α-axis and β-axis of the i-th permanent magnet motor in the two-phase stationary coordinate system, Δi αhi and Δi βhi are the high-frequency response current changes of the α-axis and β-axis of the permanent magnet motor in the two-phase stationary coordinate system, V inj represents the injected high-frequency voltage signal, θ ei represents the rotor position of the i-th permanent magnet motor, Δt represents the time change, which is two control cycles; According to the above formula of the high-frequency response current change of the permanent magnet motor, signal processing is performed and multiplied by the sign function to obtain the following formula to calculate the rotor position θ of the permanent magnet motor: ei : Among them, I sini represents the high-frequency response current envelope of the α-axis of the i-th permanent magnet motor, I cosi Represents the high-frequency response current envelope of the β-axis of the i-th permanent magnet motor.

Images