An online identification method for stator inductance of ultra-high-speed permanent magnet synchronous motor

Abstract

The present invention discloses an online identification method for the stator inductance of an ultra-high-speed permanent magnet synchronous motor, comprising the following steps: obtaining the phase current i of the motor; a and i b Perform Clark transformation to obtain the stator current i in the two-phase stationary coordinate system α and i β ; According to the stator current i in the two-phase stationary coordinate system α and i β And the voltage command u in the two-phase stationary coordinate system α and u β , obtaining sample point information; performing sample preprocessing to obtain identification sample points; and performing online stator inductance parameter identification to obtain the stator inductance value. The present invention also discloses an online identification system for the stator inductance of an ultra-high-speed permanent magnet synchronous motor. The present invention is designed based on an inductance parameter equation that does not include rotor position information. The entire identification process does not require rotor position information, which can improve the reliability of online stator inductance identification under position sensorless control conditions.

Description

Technical Field

[0001] The present invention relates to the field of motor drive control, and in particular to an online identification method for the stator inductance of an ultra-high-speed permanent magnet synchronous motor. Background Art

[0002] Permanent magnet synchronous motors (PMSMs) offer numerous advantages, including simple structure, high power density, and high operating efficiency. They are currently widely used in transportation, manufacturing, and other fields. In recent years, to further increase their power density, their design and application have shown a trend toward higher speeds, even ultra-high speeds. In specialized applications, such as air compressors in fuel cell air supply systems, the peak speed of PMSMs has exceeded 100,000 rpm. Limited by installation space and sensor sampling accuracy, ultra-high-speed (100,000 rpm and above) PMSM control systems must employ sensorless control technology to accurately and real-timely estimate the motor's rotor position online to ensure stable operation of the entire control system. Currently, most commonly used rotor position estimation methods for high-speed conditions are based on the basic voltage model of the PMSM. Therefore, the accuracy of motor parameters such as stator resistance, stator inductance, and permanent magnet flux linkage, included in the voltage model, significantly impacts the accuracy and reliability of rotor position estimation. These motor parameters can vary to varying degrees during operation due to factors such as magnetic field strength and motor temperature. Among these parameters, the stator inductance exhibits the largest and most frequent variations. To ensure accurate rotor position estimation in ultra-high-speed permanent magnet synchronous motor (PMSM) sensorless control systems, high-speed online identification of the motor's stator inductance is required. Existing online identification methods for PMSM stator inductance include the model reference adaptive observer (MRO) method and the recursive least squares method (RLSM). Most of these methods are developed for control systems with position sensors. In sensorless systems, they rely solely on estimated rotor position and speed. This inevitably introduces rotor position estimation errors into the stator inductance identification process, compromising accuracy and reliability. In severe cases, this can lead to instability in the entire control system. Summary of the Invention

[0003] The purpose of the present invention is to provide a method and system for online identification of the stator inductance of an ultra-high-speed permanent magnet synchronous motor. The present invention utilizes the characteristic that the quadrature and direct-axis inductances of ultra-high-speed motors are not much different or completely the same, thereby reducing the complexity of the calculation model and solving the problem of nonlinearity of the parameter model by using a particle swarm algorithm.

[0004] To solve the above technical problems, the present invention provides an online identification method for the stator inductance of an ultra-high-speed permanent magnet synchronous motor, comprising the following steps:

[0005] Get the phase current i of the ultra-high-speed permanent magnet synchronous motor aand i b Perform Clark transformation to obtain the stator current i in the two-phase stationary coordinate system α and i β ;

[0006] Get the voltage command u in the two-phase stationary coordinate system α and u β , according to the stator current i in the two-phase stationary coordinate system α and i β And the voltage command u in the two-phase stationary coordinate system α and u β , obtain sample point information;

[0007] Perform sample preprocessing on the sample point information to obtain the identified sample points;

[0008] The particle swarm algorithm is used to perform online parameter identification of the stator inductance on the identification sample points to obtain the stator inductance value.

[0009] Preferably, the phase current i of the ultra-high-speed permanent magnet synchronous motor is obtained a and i b , specifically including the following steps:

[0010] The switching cycle of the controller power module is a sampling time period; every sampling time period, the current sensor is used to sample and obtain the phase current i of the ultra-high-speed permanent magnet synchronous motor. a and i b .

[0011] Preferably, the voltage command u in the two-phase stationary coordinate system is obtained α and u β , according to the stator current i in the two-phase stationary coordinate system α and i β And the voltage command u in the two-phase stationary coordinate system α and u β , obtain sample point information, specifically including the following steps:

[0012] C1: According to the stator current i in the two-phase stationary coordinate system α and i β And the voltage command u in the two-phase stationary coordinate system α and u β , calculate the intermediate variable s α and s β ;

[0013]

[0014] Where: R is the stator resistance;

[0015] C2: Calculate the sample point information, which includes the two state quantities x1 and x2 of the sample point and the function value y;

[0016]

[0017] Where, ψ f is the permanent magnet flux linkage of the motor;

[0018] C3: Sample point information as S k =[x1,x2,y] is stored in the sampled sample data set; if k is equal to the preset maximum sampling number kmax, the sample point collection is completed.

[0019] Preferably, sample preprocessing is performed on the sample point information to obtain identified sample points, which specifically includes the following steps:

[0020] Calculate the sample point information S in all sampled sample data sets 1 ,S 2 ,…,S k The mean value of the state quantity and the mean of the function values

[0021] Calculate the relative error between the sample point information and the state quantity mean of each sample point, and discard the corresponding sample points whose relative error is greater than the preset maximum allowable relative error;

[0022] Count the number of remaining sample points m and determine whether the number of sample points m is less than the preset number of stator inductance online identification sample points M; if the number of sample points m is not less than the preset number of stator inductance online identification sample points M, calculate the relative error η of each remaining sample point x1 , relative error η x2 and relative error η y The sum of the errors is taken as the sum of the M sample points, and the M sample points with the smallest sum of errors are selected as the identification sample points.

[0023] Preferably, calculating the relative error between the sample point information of each sample point and the mean value of the state quantity, and discarding the corresponding sample points whose relative error is greater than a preset maximum allowable relative error, specifically includes the following steps:

[0024] Calculate the state quantity x1 and the mean value of the state quantity of the sample point information of each sample point The relative error η x1 , discard the relative error η x1 Greater than the preset maximum allowable relative error η x1,max The corresponding sample points;

[0025] Calculate the state quantity x2 and the mean value of the state quantity of the sample point information of each sample point The relative error η x2, discard the relative error η x2 Greater than the preset maximum allowable relative error η x2,max The corresponding sample points;

[0026] Calculate the function value y and the mean of the function value of the sample point information of each sample point The relative error η y , discard the relative error η y Greater than the preset maximum allowable relative error η y,max The corresponding sample points.

[0027] Preferably, the particle swarm algorithm is used to perform online parameter identification of the stator inductance on the identification sample points to obtain the stator inductance value, which specifically includes the following steps:

[0028] E1: Initialize the particle swarm, randomly generate N-2 particles within the possible range of stator inductance, add the initial value of stator inductance and the stator inductance value identified last time, and the total number of N particles constitutes the initial position of the first generation particle swarm, recorded as At the same time, the initial velocity of each particle in the first generation population is generated within the possible range of the stator inductance change speed, which is recorded as

[0029] E2: Use the identified sample points to calculate the fitness value J of each particle, and take the initial position of each particle as its current individual optimal position, recorded as P1, P2…P N , the position with the lowest fitness value J is taken as the current optimal position of the population, denoted as G best , the fitness value is calculated as follows;

[0030]

[0031] Where: is the position of the particle; J is the fitness value; Represents the y value calculated using the current particle position, and j represents the jth sampling point;

[0032] E3: Update the speed and position of each particle in the population according to the following formula to obtain the new particle position and speed;

[0033]

[0034] In the formula is the inertia weight coefficient that decreases with the number of iterations, c1 and c2 are learning factors, generally positive numbers less than 1, r1 and r2 are random numbers between 0 and 1; h is the number of generations of the population.

[0035] E4: Calculate the fitness value of the new particle position;

[0036] If the fitness value of the new particle position is less than the fitness value of the current individual optimal position of the particle, the new particle position will be used as the individual optimal position of the particle;

[0037] If the fitness value of the new particle position is less than the current optimal position of the particle population G best If the fitness value is , the new particle position is taken as the current optimal position of the population;

[0038] E5; Determine whether the fitness value J of the current optimal position of the population is less than the preset iteration termination condition;

[0039] If the fitness value J of the current population optimal position is less than the preset iteration termination condition, the current population optimal position is used as the stator inductance value finally estimated by the particle swarm algorithm; otherwise, E3 is re-executed.

[0040] The present invention also provides a system for implementing the above-mentioned method for online identification of stator inductance of an ultra-high-speed permanent magnet synchronous motor, comprising:

[0041] Clark transformation module is used to transform the phase current i a and i b Perform Clark transformation to obtain the stator current i in the two-phase stationary coordinate system α and i β ;

[0042] The sample preprocessing module is used to calculate the stator current i in the two-phase stationary coordinate system. α and i β And the voltage command u in the two-phase stationary coordinate system α and u β , obtain sample point information; perform sample preprocessing on the sample points to obtain identified sample points;

[0043] The parameter identification module is used to perform online parameter identification of the stator inductance on the identification sample points using the particle swarm algorithm to obtain the stator inductance value.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] First, the stator inductance online identification method described in the present invention is designed based on an inductance parameter equation that does not include rotor position information. The entire identification process does not require rotor position information, which can improve the reliability of stator inductance online identification under position sensorless control conditions.

[0046] Second, the online stator inductance identification method of the present invention utilizes the characteristic that the quadrature and direct axis inductances of ultra-high-speed motors are similar or identical, reducing the complexity of the computational model. It also uses a particle swarm optimization algorithm to address the nonlinearity of the parameter model.

[0047] Third, the present invention preprocesses the sample points used by the particle swarm optimization algorithm, eliminating sample points with low confidence levels. At the same time, it effectively controls the total number of sample points and reduces the computational complexity of the particle swarm algorithm, ensuring that it can complete the computational process in a relatively short time and ensuring real-time performance. Fourth, the present invention utilizes the variation pattern of the stator inductance to optimize the initialization process of the particle swarm optimization algorithm. On the basis of the original purely randomly generated initial particles, two initial particles, the initial value and the stator inductance value identified last time, are added to improve the convergence speed of parameter identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 This is a structural block diagram of the ultra-high-speed permanent magnet synchronous motor stator inductance online identification system of the present invention;

[0049] Figure 2 Flowchart of stator inductance online identification sample point collection and preprocessing;

[0050] Figure 3 Flowchart of online identification of stator inductance based on particle swarm optimization algorithm;

[0051] Figure 4 This is a schematic diagram of the overall processing flow of the online identification system for the stator inductance of an ultra-high-speed permanent magnet synchronous motor of the present invention; DETAILED DESCRIPTION

[0052] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0053] The present invention will be described in further detail below with reference to the accompanying drawings:

[0054] The present invention discloses an online identification method for the stator inductance of an ultra-high-speed permanent magnet synchronous motor, comprising the following steps:

[0055] Based on the classic current closed-loop control framework of permanent magnet synchronous motor, the sampled phase current of ultra-high-speed permanent magnet synchronous motor is sent to the sample preprocessing module together with the phase voltage command after coordinate transformation for sample preprocessing. The sample points after preprocessing are sent to the stator inductance online identification module based on particle swarm algorithm. The identified stator inductance will be used in real time in the rotor position observer of the ultra-high-speed permanent magnet synchronous motor position sensorless control system.

[0056] The present invention is different from most permanent magnet synchronous motor parameter identification methods. It no longer performs parameter identification based on the basic voltage equation of the permanent magnet synchronous motor. Instead, it takes advantage of the fact that most ultra-high-speed permanent magnet synchronous motors are hidden-pole motors. Assuming that the quadrature and direct-axis inductances of the motor are equal, the stator inductance parameter equation that does not contain rotor position information is derived based on the motor flux equation. This equation is then used to realize online identification of the stator inductance without a position sensor. Based on the known stator resistance and permanent magnet flux of the motor, this method uses the voltage and current information of the permanent magnet synchronous motor that can be directly obtained as input signals, and uses a particle swarm algorithm to perform online identification of the motor stator inductance. At the same time, the population initialization process of the particle swarm algorithm is optimized based on the changing characteristics of the stator inductance, and the sample preprocessing process of the input signal is combined to reduce the amount of computation required for online identification and improve its computational speed.

[0057] In order to better illustrate the technical effects of the present invention, the present invention provides the following specific embodiments to illustrate the above technical process, such as Figure 1-4 The method specifically includes the following steps:

[0058] A: The speed of the ultra-high-speed permanent magnet synchronous motor is controlled according to the classic closed-loop vector control theory based on the motor speed command. The motor rotor position required for vector control can be obtained using any model-based rotor position observer.

[0059] The switching cycle of the controller power module is a sampling time period; the phase current i of the ultra-high-speed permanent magnet synchronous motor is obtained by sampling with a current sensor every other sampling time period. a and i b ;

[0060] B: Phase current i a and i b Send it to the Clark transformation module for Clark transformation, and convert the phase current i a and i b Converted to the stator current i in the two-phase stationary coordinate system α and i β ;

[0061] C: Get the voltage command u in the two-phase stationary coordinate system α and u β , the stator current i in the two-phase stationary coordinate system α and i β And the voltage command u in the two-phase stationary coordinate system α and u β The sample is sent to the sample preprocessing module, and the motor electrical angular velocity observed by the rotor position observer is sent to the sample preprocessing module. Send to the sample preprocessing module to obtain sample point information and then perform sample preprocessing, such as Figure 2As shown, the specific steps include:

[0062] C1: The intermediate variable s is calculated using the input value and stator resistance R according to the following formula α and s β ;

[0063]

[0064] C2: Calculates the sample point information according to the following formula. The sample point information includes the two state quantities x1 and x2 of the sample point and the function value y.

[0065]

[0066] Where, ψ f is the permanent magnet flux linkage of the motor;

[0067] C3: The sample point information S calculated in step C2 k =[x1,x2,y] is stored in the sampled data set;

[0068] Where: S k is the sample point information of the kth sample point.

[0069] C4: Determine whether k is equal to the preset maximum sampling number kmax;

[0070] If k is equal to the preset maximum sampling number kmax, execute step C5; otherwise, execute step C6;

[0071] C5: Sample collection is complete and pre-processing is performed. The specific process includes:

[0072] C5-1: Calculate the sample point information S in all sampled sample data sets 1 ,S 2 ,…,S k The mean value of the state quantity and the mean of the function values

[0073] C5-2: Calculate the state quantity x1 and the mean of the state quantity of the sample point information of each sample point The relative error η x1 , discard the relative error η x1 Greater than the preset maximum allowable relative error η x1,max The corresponding sample points;

[0074] C5-3: Calculate the state quantity x2 and the mean of the state quantity of the sample point information of each sample point The relative error η x2 , discard the relative error η x2 Greater than the preset maximum allowable relative error ηx2,max The corresponding sample points;

[0075] C5-4: Calculate the function value y and the mean of the function value of the sample point information of each sample point The relative error η y , discard the relative error η y Greater than the preset maximum allowable relative error η y,max The corresponding sample points;

[0076] C5-5: Count the number of remaining sample points m and determine whether the number of sample points m is less than the preset number of stator inductance online identification sample points M;

[0077] If the number of sample points m is less than the preset number of stator inductance online identification sample points M, then execute step C5-6; otherwise, execute step C5-7;

[0078] C5-6: This group of sampling points is invalid, clear all sampling points, set k = 1, restart sampling in the next sampling cycle, and re-execute step A;

[0079] C5-7: For each remaining sample point, the sum of the relative errors of the state quantity x1, state quantity x2 and function value y is calculated (i.e., the relative error η x1 , relative error η x2 , relative error η y ), select M sample points with the smallest sum of errors as identification sample points for online identification of stator inductance, set k = 1, and set the identification enable signal to 1; execute step D;

[0080] C6: Sample point collection is not completed, add 1 to k, continue sampling in the next sampling cycle, and execute step A again;

[0081] D: Determine whether the identification enable signal is 1. If it is 1, execute E; if not, execute F.

[0082] E: Based on the identification sample points obtained from C5-7, the particle swarm algorithm is used to perform online parameter identification of the stator inductance to obtain the stator inductance value, such as Figure 3 As shown, the specific steps include:

[0083] E1: Initialize the particle swarm, randomly generate N-2 particles within the possible range of stator inductance, add the initial value of stator inductance and the stator inductance value identified last time, and the total number of N particles constitutes the initial position of the first generation particle swarm, recorded as At the same time, the initial velocity of each particle in the first generation population is generated within the possible range of the stator inductance change speed, which is recorded as

[0084] as well as The superscript 1 in represents the number of generations of the population;

[0085] In this embodiment, the stator inductance may have a value range of 0.2 to 2 times the initial value of the stator inductance, and the change speed may have a value range of 2 times the initial value of the stator inductance / ms.

[0086] E2: Use the identified sample points to calculate the fitness value J of each particle, and take the initial position of each particle as its current individual optimal position, recorded as P1, P2…P N , the position with the lowest fitness value J is taken as the current optimal position of the population, denoted as G best , the fitness value is calculated as follows;

[0087]

[0088] Where: is the position of the particle; J is the fitness value; Represents the y value calculated using the current particle position, j represents the jth sampling point, and there are M total.

[0089] E3: Update the speed and position of each particle in the population according to the following formula to obtain the new particle position and speed;

[0090]

[0091] Where: is the inertia weight coefficient that decreases with the number of iterations, c1 and c2 are learning factors, generally positive numbers less than 1, r1 and r2 are random numbers between 0 and 1; h is the number of generations of the population.

[0092] E4: Calculate the fitness value of the new particle position;

[0093] If the fitness value of the new particle position is less than the fitness value of the current individual optimal position of the particle, the new particle position will be used as the individual optimal position of the particle;

[0094] If the fitness value of the new particle position is less than the fitness value of the current optimal position of the particle, the new particle position will be used as the current optimal position of the population;

[0095] E5; Determine whether the fitness value J of the current optimal position of the population is less than the preset iteration termination condition, that is, whether it is less than a sufficiently small value;

[0096] If the fitness value J of the current optimal position of the population is less than the preset iteration termination condition, execute E6; otherwise, execute E3 again;

[0097] E6: Use the current optimal position of the population as the final estimated stator inductance value obtained by the particle swarm algorithm, and set the identification enable signal to 0;

[0098] F: The stator inductance value is sent to the rotor position observer. The stator inductance value is used for rotor position and speed observation without position sensor.

[0099] The present invention also provides a system for implementing the above-mentioned method for online identification of the stator inductance of an ultra-high-speed permanent magnet synchronous motor, comprising:

[0100] Clark transformation module; has the phase current i a and i b Perform Clark transformation to obtain the stator current i in the two-phase stationary coordinate system α and i β ;

[0101] The sample preprocessing module is used to calculate the stator current i in the two-phase stationary coordinate system. α and i β And the voltage command u in the two-phase stationary coordinate system α and u β Sampling is performed to obtain sample points; sample preprocessing is performed on the sample points to obtain identification sample points;

[0102] The parameter identification module is used to perform online parameter identification of the stator inductance on the identification sample points using the particle swarm algorithm to obtain the stator inductance value.

[0103] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions within the technical scope disclosed by the present invention shall be covered by the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be subject to the scope of protection of the claims.

Claims

1. A method for online identification of stator inductance of an ultra-high-speed permanent magnet synchronous motor, characterized in that: The following steps are involved: Obtaining Phase Currents of Ultra-High-Speed ​​Permanent Magnet Synchronous Motors and Perform Clark transformation to obtain the stator current in the two-phase stationary coordinate system and ; According to the stator current in the two-phase stationary coordinate system and And the voltage command in the two-phase stationary coordinate system and , calculate the intermediate variable and ; Where: R is the stator resistance; Calculate the sample point information, which includes the two state quantities x1, x2 and the function value y of the sample point; Where, is the permanent magnet flux linkage of the motor; is the motor electrical angular velocity; Sample point information as Store it in the sampled sample data set; if k is equal to the preset maximum sampling number kmax, the sample point collection is completed; Perform sample preprocessing on the sample point information to obtain the identified sample points; The particle swarm algorithm is used to perform online parameter identification of the stator inductance on the identification sample points to obtain the stator inductance value.

2. The method for online identification of stator inductance of an ultra-high-speed permanent magnet synchronous motor according to claim 1, characterized in that: Obtaining Phase Currents of Ultra-High-Speed ​​Permanent Magnet Synchronous Motors and , specifically including the following steps: The switching cycle of the controller power module is used as a sampling time period; every other sampling time period, the current sensor is used to sample and obtain the phase current of the ultra-high-speed permanent magnet synchronous motor. .

3. The method for online identification of stator inductance of an ultra-high-speed permanent magnet synchronous motor according to claim 1, characterized in that: The sample point information is pre-processed to obtain the identified sample points, which specifically includes the following steps: Calculate the sample point information of all sampled sample data sets The mean value of the state quantity and the mean of the function values ; Calculate the relative error between the sample point information and the state quantity mean of each sample point, and discard the corresponding sample points whose relative error is greater than the preset maximum allowable relative error; Count the number of remaining sample points m and determine whether the number of sample points m is less than the preset number of stator inductance online identification sample points M; if the number of sample points m is not less than the preset number of stator inductance online identification sample points M, calculate the relative error of each remaining sample point separately. , relative error and relative error The sum of the errors is taken as the sum of the M sample points, and the M sample points with the smallest sum of errors are selected as the identification sample points.

4. The method for online identification of stator inductance of an ultra-high-speed permanent magnet synchronous motor according to claim 3, characterized in that: Calculate the relative error between the sample point information and the state quantity mean of each sample point, and discard the corresponding sample points whose relative error is greater than the preset maximum allowable relative error. Specifically, the following steps are included: Calculate the state quantity of the sample point information of each sample point and the mean of state quantities The relative error , discard the relative error Greater than the preset maximum allowable relative error The corresponding sample points of Calculate the state quantity of the sample point information of each sample point and the mean of state quantities The relative error , discard the relative error Greater than the preset maximum allowable relative error The corresponding sample points of Calculate the function value y and the mean of the function value of the sample point information of each sample point The relative error , discard the relative error Greater than the preset maximum allowable relative error The corresponding sample points.

5. The method for online identification of stator inductance of an ultra-high-speed permanent magnet synchronous motor according to claim 4, characterized in that: The particle swarm algorithm is used to perform online parameter identification of the stator inductance on the identification sample points to obtain the stator inductance value. The specific steps include: E1: Initialize the particle swarm, randomly generate N-2 particles within the possible range of stator inductance, add the initial value of stator inductance and the stator inductance value identified last time, and the total number of N particles constitutes the initial position of the first generation particle swarm, recorded as At the same time, the initial velocity of each particle in the first generation population is generated within the possible range of the stator inductance change speed, which is recorded as ; E2: Use the identified sample points to calculate the fitness value J of each particle, and take the initial position of each particle as its current individual optimal position, recorded as , the position with the lowest fitness value J is taken as the current optimal position of the population, recorded as , the fitness value is calculated as follows; Where: is the position of the particle; J is the fitness value; represents the y value calculated using the current particle position, and j represents the jth sampling point; E3: Update the speed and position of each particle in the population according to the following formula to obtain the new particle position and speed; In the formula is the inertia weight coefficient that decreases as the number of iterations increases, and is the learning factor, and is a positive number less than 1, and is a random number between 0 and 1; is the algebra of the population; E4: Calculate the fitness value of the new particle position; If the fitness value of the new particle position is less than the fitness value of the current individual optimal position of the particle, the new particle position will be used as the individual optimal position of the particle; If the fitness value of the new particle position is less than the current optimal position of the particle population If the fitness value is , the new particle position is taken as the current optimal position of the population; E5; determine whether the fitness value J of the current population optimal position is less than the preset iteration termination condition; If the fitness value J of the current population's optimal position is less than the preset iteration termination condition, the current population's optimal position The stator inductance value obtained as the final estimation by the particle swarm algorithm; otherwise, re-execute E3.

6. A system for implementing the method for online identification of stator inductance of an ultra-high-speed permanent magnet synchronous motor according to any one of claims 1 to 5, characterized in that: include: Clark transformation module for phase current Perform Clark transformation to obtain the stator current in the two-phase stationary coordinate system ; Sample preprocessing module, used to calculate the stator current in the two-phase stationary coordinate system And the voltage command in the two-phase stationary coordinate system , obtain sample point information; Perform sample preprocessing on the sample points to obtain identification sample points; The parameter identification module is used to perform online parameter identification of the stator inductance on the identification sample points using the particle swarm algorithm to obtain the stator inductance value.

Images