Method, device and storage medium for identifying motor parameters

Abstract

The application relates to a motor parameter identification method and device and a storage medium, wherein the method comprises the following steps: S11, initializing parameters of a parrot algorithm through a preset sequence function, and generating corresponding parrot populations based on the initialized parrot algorithm; S12, determining individual fitness of each parrot in the parrot populations, wherein the individual fitness represents the distance between the parrot and a target position; S13, determining a target parrot of an optimal position in the parrot populations according to the individual fitness, and updating the position of the parrot according to a target behavior of the parrot, wherein the target behavior represents the behavior of the parrot approaching the target parrot; and S14, repeatedly executing S12 and S13 until a performance index meets a preset condition to determine a motor parameter corresponding to a position vector. Through the application, the problem that the accuracy of motor parameter identification is low and the identification process is complex in the related art is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of motor parameter identification, and in particular to a motor parameter identification method, device and storage medium. BACKGROUND

[0002] In the existing motor control system, the output efficiency of the motor, the arrangement of the internal system circuit of the motor, and the setting of the parameters of the controllers of the motor are greatly dependent on accurate motor parameters. Considering that the motor is a kind of electronic device that is extremely susceptible to environmental interference, temperature changes, air humidity, etc., the above weather changes will inevitably cause the motor parameters to change.

[0003] The commonly used motor parameter identification strategies currently include: 1) recursive least squares method, which is applied to motor parameter identification. This method needs to constantly learn old information in identification, which is prone to data saturation. 2) Combining model reference adaptive method with motor parameter identification, but this method has the defects of complex adjustable model establishment and subjective experience-dependent adaptive rate selection. 3) Applying Kalman filter algorithm to motor parameter identification. This method estimates the state at the next time according to the estimated value and covariance at the present time, but the related parameters of the Kalman gain matrix of this algorithm are not easy to select, and a large number of matrix operations are involved in solving, which is not conducive to engineering use. It can be seen that the accuracy of motor parameter identification in the related technology is low and the identification process is complex.

[0004] In view of the above technical problems in the related art, there is currently no effective solution. SUMMARY

[0005] The present application provides a motor parameter identification method, device and storage medium to solve the problem of low accuracy of motor parameter identification and complex identification process in the related art.

[0006] In a first aspect, the present application provides a motor parameter identification method, comprising:

[0007] S11, initializing the parameters of the parrot algorithm by a preset sequence function, and generating a corresponding parrot population based on the initialized parrot algorithm; wherein the parrot population includes a plurality of parrots, the parrot is a position vector of a motor parameter to be identified, the dimension of the position vector represents the number of the motor parameter to be identified, and the motor parameter to be identified is a motor parameter in a motor double closed loop model;

[0008] S12, determining the individual fitness of each parrot in the parrot population, wherein the individual fitness represents the distance of the parrot from the target position;

[0009] S13, determining a target parrot in an optimal position in the parrot population according to the individual fitness, and updating the position of the parrot according to a target behavior of the parrot, wherein the target behavior represents the behavior of the parrot approaching the target parrot;

[0010] S14, repeating S12 and S13 until a performance index meets a preset condition to determine the motor parameter corresponding to the position vector, wherein the performance index is a performance index of the motor double closed loop model determined based on the parrot after the position is updated.

[0011] Optionally, the parameters of the parrot algorithm are initialized through a preset sequence, including initializing the parameters of the parrot algorithm through a Halton sequence.

[0012] The Halton sequence is H(n) is the initial position of the nth parrot, n is the ordinal number of the Halton sequence, p is the base of the Halton sequence, and is called the root inverse of the initial position n of the parrot with respect to the base p1, and p i (1≤i≤s), wherein s is the number of to-be-identified parameters, a i is a constant variable.

[0013] Optionally, when the target behavior is foraging behavior, updating the position of the parrot according to the target behavior of the parrot includes: determining a flight strategy of the parrot through a first formula, and updating the position of the parrot in the foraging process according to a second formula under the flight strategy.

[0014] The first formula is:

[0015]

[0016] wherein, is the position vector of the t+1 generation parrot, σ represents the flight distance of the parrot, and tan(θ) is the direction of the flight of the parrot.

[0017] The second formula is:

[0018]

[0019] wherein, is the mean value of the position in the t generation parrot population, max is the distance of the parrot in the parrot population farthest from the target position in this round of iteration, Levy(d) represents a function of the flight of the parrot, and x best is the position of the leading parrot, and rand represents any number in the range of ​

[0020] Optionally, in the case where the target behavior is a stay behavior, updating the position of the parrot according to the target behavior of the parrot comprises: determining a flight strategy of the parrot by a first formula, and updating the position of the parrot in the stay process according to a third formula under the flight strategy;

[0021] The first formula is:

[0022]

[0023] wherein, is the position vector of the t+1 generation parrot, σ represents the flight distance of the parrot, and tan(θ) is the direction of the flight of the parrot;

[0024] The third formula is:

[0025]

[0026] wherein, ones(1, d) represents a vector composed of 1 and having a dimension of d, x best is the position of the current leader parrot, and Levy(d) represents a function representing the flight of the parrot.

[0027] Optionally, in the case where the target behavior is a communication behavior, updating the position of the parrot according to the target behavior of the parrot comprises: updating the position of the parrot in the communication behavior by a fourth formula;

[0028] The fourth formula is:

[0029]

[0030] wherein, p is a probability control variable.

[0031] Optionally, in the case where the target behavior is a fear behavior, updating the position of the parrot according to the target behavior of the parrot comprises:

[0032] updating the position of the parrot in the fear by a fifth formula;

[0033]

[0034] Optionally, S2 and S3 are repeatedly executed until the motor parameters corresponding to the position vector are determined when the performance index meets the preset condition, comprising:

[0035] S21, dividing the parrots after the update of the positions of the parrots obtained by performing S12 and S13 once into parrots of a first preset proportion and parrots of a second preset proportion, wherein the sum of the first preset proportion and the second preset proportion is 1, and the first preset proportion is smaller than the second preset proportion;

[0036] S22, updating the positions of the parrots of the first preset proportion by a sixth formula;

[0037] wherein the sixth formula is:

[0038] x s (t+1)=x s (t)+S(r1x best -r2x s (t))

[0039] x s represents the position of the t+1 generation parrot of the s-th parrot; x best is the parrot closest to the target position in the parrot group; r is a random number (0-1); and S is a flip factor;

[0040] S=Ks+ε*ΔK s

[0041] K s is an uncorrected flip factor, ε represents a proportion factor, and ΔK s represents a flip factor gain after the correction by the fuzzy control;

[0042] S23, updating the individual fitness of the parrots of the first preset proportion and the parrots of the second preset proportion after the update of the positions, and performing S12 again according to the updated individual fitness to re-determine the target parrot;

[0043] S24, performing the update operation of the position of the parrot according to the target behavior of the parrot in S13 again until the performance index meets the preset condition to determine the motor parameter corresponding to the position vector.

[0044] Optionally, the performance index is determined by: determining the integral of the absolute value of the product of the time t and the speed loop error in the motor double closed loop model; and taking the sum of the integral and the product of the time and the current regulation loop error as the performance index.

[0045] In a second aspect, the present application provides a device, comprising: at least one communication interface; at least one bus connected with the at least one communication interface; at least one processor connected with the at least one bus; and at least one memory connected with the at least one bus, wherein the processor is configured to execute the motor parameter identification method in the first aspect of the present application.

[0046] In a third aspect, the present application also provides a computer storage medium storing computer executable instructions for executing the motor parameter identification method of the first aspect of the present application.

[0047] Compared with the prior art, the above technical solution provided by the embodiments of the present application has the following advantages: in the method provided by the embodiments of the present application, the parameters of the parrot algorithm are initialized by a preset sequence function, and the corresponding parrot population is generated based on the initialized parrot algorithm, that is, the motor parameters to be identified in the motor double closed-loop model are mapped to parrots, then the individual fitness of the parrot is determined, that is, the distance of the parrot from the target position, then the target parrot of the optimal position in the parrot population is determined through the individual fitness, and the position of the parrot is updated according to the target behavior of the parrot. The updating process is a continuous iteration process, which can continuously update the position of the parrot, improve the identification efficiency of the motor parameters, and improve the identification accuracy of the motor parameters, thereby solving the problems of low accuracy and complex identification process of motor parameter identification in the related art. BRIEF DESCRIPTION OF DRAWINGS

[0048] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and serve to explain the principles of the present application, together with the description.

[0049] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the drawings needed in the embodiments or the prior art description will be briefly introduced as follows. Obviously, for those skilled in the art, other drawings can also be obtained without creative labor based on these drawings.

[0050] One or more embodiments are exemplarily illustrated by the pictures in the corresponding drawings, which do not constitute a limitation on the embodiments. Elements with the same reference numerals in the drawings represent similar elements, unless otherwise specified. The drawings in the drawings do not constitute a proportional limitation.

[0051] Figure 1 A flowchart of a motor parameter identification method provided by the embodiments of the present application;

[0052] Figure 2 A schematic diagram of a motor double closed-loop model provided by the embodiments of the present application;

[0053] Figure 3 A schematic diagram of a motor double closed-loop model with performance indicators provided by the embodiments of the present application;

[0054] Figure 4A flowchart of a parrot algorithm parameter identification method provided by an embodiment of the present application is shown in the figure;

[0055] Figure 5 A schematic diagram of a fuzzy surface of ΔKs provided by an embodiment of the present application is shown in the figure;

[0056] Figure 6 A structural schematic diagram of the device provided by an embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0057] To make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described below in connection with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0058] The following disclosure provides many different embodiments, or examples, for implementing different structures of the present application. For the purpose of simplicity, the components and arrangements of the examples that are described herein are shown in the following description. These are, of course, merely examples and are not intended to limit the application. Moreover, the application can be practiced with different examples than those described herein without departing from the scope of the application. Further, the described embodiments are related to the best mode contemplated for carrying out the application and comprises the combinations of elements and / or features described herein that are intended to provide advantages for the practicing the application. The elements and / or features of described embodiments can be combined in other combinations than those described to provide advantages for the practicing the application.

[0059] To solve the problem of low accuracy and complex identification process of motor parameter identification in the related art, the present application provides a motor parameter identification method, as shown in Figure 1 The steps of the method include:

[0060] S11, initializing parameters of a parrot algorithm through a preset sequence function, and generating a corresponding parrot population based on the initialized parrot algorithm; wherein the parrot population includes a plurality of parrots, the parrot is a position vector of a motor parameter to be identified, the dimension of the position vector represents the number of the motor parameter to be identified, and the motor parameter to be identified is a motor parameter in a motor double closed loop model;

[0061] S12, determining individual fitness of each parrot in the parrot population, wherein the individual fitness represents the distance of the parrot from the target position;

[0062] S13, determining a target parrot of an optimal position in the parrot population according to the individual fitness, and updating the position of the parrot according to the target behavior of the parrot, wherein the target behavior represents the behavior of the parrot approaching the target parrot;

[0063] S14, repeating S12 and S13 until the motor parameters corresponding to the position vector are determined when the performance index meets the preset condition, wherein the performance index refers to the performance index of the motor double closed loop model determined by the parrot based on the updated position.

[0064] Through steps S11 to S14, the parameters of the parrot algorithm are initialized by the preset sequence function, and the corresponding parrot population is generated based on the initialized parrot algorithm, that is, the to-be-identified motor parameters in the motor double closed loop model are mapped to the parrot, then the individual fitness of the parrot, that is, the distance between the parrot and the target position, is determined, then the target parrot of the optimal position in the parrot population is determined through the individual fitness, and the position of the parrot is updated according to the target behavior of the parrot. The updating process is a continuous iteration process, which can continuously update the position of the parrot, improve the identification efficiency of the motor parameters, and improve the identification accuracy of the motor parameters, thereby solving the problems of low identification accuracy and complex identification process of the motor parameters in the related art.

[0065] In the embodiments of the present application, the way of repeatedly performing S2 and S3 until the motor parameters corresponding to the position vector are determined when the performance index meets the preset condition in step 14 can further include:

[0066] S21, dividing the parrots after the updated position obtained by performing S12 and S13 once into parrots of a first preset proportion and parrots of a second preset proportion, wherein the sum of the first preset proportion and the second preset proportion is 1, and the first preset proportion is less than the second preset proportion;

[0067] S22, updating the position of the parrots of the first preset proportion by the sixth formula;

[0068] Wherein the sixth formula is:

[0069] x s (t+1)=x s (t)+S(r1x best -r2x s (t))

[0070] x s represents the t+1 generation parrot position of the s-th parrot; x best is the parrot closest to the target position in the parrot population; r is a random number (0-1); S is a flip factor;

[0071] S=Ks+ε*ΔK s

[0072] K s is the uncorrected flip factor, ε represents the proportion factor, and ΔK sa representative of the roll factor gain after being corrected by the fuzzy control;

[0073] S23, updating the individual fitness of the first preset proportion of parrots and the second preset proportion of parrots after the position update, and performing S12 again according to the updated individual fitness to re-determine the target parrot;

[0074] S24, performing the updating operation of the position of the parrot according to the target behavior of the parrot in S13 again until the performance index meets the preset condition to determine the motor parameter corresponding to the position vector.

[0075] In specific examples, the first preset proportion can be 0.1, and the second preset proportion can be 0.9. The above is only an example, and can be set according to the needs in actual application scenarios.

[0076] Through the above steps S21 and S24, the individual fitness of the parrot can be continuously updated and iterated. Since the behavior of the parrot changes its distance to the target position every time, the optimal position, i.e. the parrot closest to the target, needs to be selected again after each update iteration. The iteration is repeated until the performance index of the parrot identified by the motor parameter after the iteration meets the preset condition, so that the iteration can be terminated.

[0077] In the embodiments of the present application, the method of initializing the parameters of the parrot algorithm through the preset sequence in step S11 can be: initializing the parameters of the parrot algorithm through a Halton sequence.

[0078] Wherein, the Halton sequence is H(n) is the initial position of the nth parrot, n is the ordinal number of the Halton sequence, p is the base of the Halton sequence, and is called the root inverse of the initial position n of the parrot with respect to the base p1, and p i (1≤i≤s), where s is the number of parameters to be identified, a i is a constant variable.

[0079] The existing parrot population initialization method depends on a rand function to generate a random sequence, and the rand function has the advantage of simple population initialization process, but the generation of the rand random sequence is easy to cause uneven distribution of the parrot population, which can cause blank in the search space and reduce the efficiency of the parrot search. In the embodiment of the application, the Halton sequence is considered to be a small difference sequence, which is constructed by a prime number as a base, that is, the Halton sequence is used to replace the existing rand sequence in the embodiment of the application, so that the parrot population is densely distributed in the search space, thereby improving the optimization efficiency of the parrot algorithm.

[0080] In the embodiment of the application, the parrots in the parrot population have various behaviors, such as foraging behavior, staying behavior, communication behavior and fear behavior. The position updating mode of the parrot will be explained in combination with each specific behavior. In the case of the target behavior being foraging behavior, the way of updating the position of the parrot according to the target behavior of the parrot in step 13 can further include:

[0081] Step 31, determining the flight strategy of the parrot by the first formula, and updating the position of the parrot in the foraging process according to the second formula under the flight strategy;

[0082] The first formula is:

[0083]

[0084] Wherein, is the position vector of the t+1 generation parrot, and σ represents the flight distance of the parrot; tan(θ) is the direction of the flight of the parrot;

[0085] The second formula is:

[0086]

[0087] Wherein, is the mean value of the position in the t generation parrot population, max is the distance of the parrot population farthest from the target position in the parrot population in this round of iteration; Levy(d) represents a function of the flight of the parrot; x best is the position of the leader parrot, that is, the position of the best parrot in this round of iteration, and rand represents any number in the range of

[0088] In the embodiment of the application, in the case of the target behavior being staying behavior, the way of updating the position of the parrot according to the target behavior of the parrot in step 13 can further include:

[0089] ​Step 41, determining the flight strategy of the parrot by the first formula, and updating the position of the parrot in the staying process according to the third formula under the flight strategy;

[0090] The first formula is:

[0091]

[0092] wherein, is the position vector of the t+1 generation parrot, σ represents the flight distance of the parrot, and tan(θ) is the direction of the flight of the parrot;

[0093] The third formula is:

[0094]

[0095] wherein, ones(1, d) represents a vector composed of 1 and having a dimension of d, x best is the position of the current leader parrot, and Levy(d) represents a function of the flight of the parrot.

[0096] In the embodiment of the present application, in the case where the target behavior is the communication behavior, the manner of updating the position of the parrot according to the target behavior of the parrot in the above step 13 can further include:

[0097] Step 51, updating the position of the parrot in the communication behavior by the fourth formula;

[0098] The fourth formula is:

[0099]

[0100] wherein, p is a probability control variable.

[0101] In the embodiment of the present application, in the case where the target behavior is the fear behavior, the manner of updating the position of the parrot according to the target behavior of the parrot in the above step 13 can further include:

[0102] Step 61, updating the position of the parrot in the fear by the fifth formula;

[0103]

[0104] In the embodiment of the present application, the performance index is determined by the following manner: first, determining the integral of the product absolute value of the time t and the speed loop error in the motor double closed loop model, and then taking the sum of the integral, the time, and the product absolute value of the current regulation loop error as the performance index.

[0105] In a specific example, the performance index can be the integral of the absolute value of the product of time t and the permanent magnet synchronous motor system speed loop error E(t) and the sum of the absolute value of the product of t and the current regulation loop error e(t). The performance index quantifies the performance of the motor parameters solved based on the parrot algorithm and returned by calling the motor system model. The mathematical expression of the performance index J is as follows:

[0106]

[0107] In the embodiments of the present application, the synchronous motor double closed loop system continuously generates the speed loop error E(t) and the current loop error e(t). When the motor system is running, the changes of E(t) and e(t) are detected at any time, so as to adaptively generate a series of different sizes of J. The smaller the value is, the smaller the deviation of the speed loop and the current loop of the motor system is, that is, the system is in a steady state at this time, and the identified motor parameters are more accurate.

[0108] The embodiments of the present application are explained and described below. The specific embodiments provide a permanent magnet synchronous motor parameter identification method based on an improved parrot algorithm. The existing permanent magnet synchronous motor double loop control model, as shown in Figure 2 , when the parrot algorithm is deployed, a performance index function needs to be added to the permanent magnet synchronous motor double loop control model, as shown in the dashed box in Figure 3 .

[0109] The embodiments are described below. In Figure 2 , the specific embodiments adopt a double closed loop motor control system, which sets the given d-axis current to zero as the control scheme of the whole system. Specifically, the difference between the given speed and the speed feedback is obtained as E(t). The given q-axis input current is obtained through the speed regulation loop. At this time, the current deviation e(t) formed by the difference between the given d-axis and q-axis currents and the corresponding feedback currents is taken as the input of the current regulation loop, so as to generate the given input voltage of the q-axis and d-axis, respectively. Then, through 2 / 2 transformation, the voltage in the d-axis and q-axis coordinate system is converted into the voltage corresponding to the alpha and beta coordinate system. The voltage generates the switching signal of the IGBT through the action of the SVPWM module, so as to realize the three-phase current I a , I b , I c , and the electrical angle θ of the output signal of the motor. The above output quantities of the motor, the current and the electrical angle, are converted through 3 / 2 and 2 / 2 coordinate systems to generate the d-axis and q-axis current feedback. The difference between the feedback and the given d-axis and q-axis currents generates E(t), so as to realize closed loop control.

[0110] The motor double closed loop model after adding the performance index is as shown in Figure 3the dashed box in the figure, which is specifically presented by formula (10) in the specific embodiment, the input of the dashed box is the deviation amount E(t) between the given rotating speed and the feedback rotating speed, through multiplication operation with time and then integral operation, finally summing the value obtained by multiplying the d-axis current deviation amount e(t) with time and then integrating, the "output" of the performance index part is finally obtained, which corresponds to Figure 4 the "calculating and returning performance index" step in the right dashed box. In combination with Figure 4 It can be known that the method for identifying motor parameters by the parrot algorithm in the embodiment of the present application includes:

[0111] Step 401, parameter initialization;

[0112] The existing parrot population initialization function is:

[0113] x i =rand(ub-lb)+lb

[0114] wherein, ub is the upper bound of the range containing the set of motor parameters to be identified; lb is the lower bound of the range containing the set of motor parameters to be identified; rand represents an arbitrary number in the range of [0, 1]; x i is the initial position of the i-th parrot in foraging.

[0115] However, the existing parrot population initialization method excessively relies on the rand function to generate a random sequence, and the rand function has the advantage of simple population initialization process, but the generation of the rand random sequence is prone to the problem of uneven distribution of the parrot population, which in turn causes the formation of blank in the search space and reduces the efficiency of parrot search. In the embodiment of the present application, the Halton sequence is used, which is a kind of small difference sequence constructed by a prime number as a base, that is, the Halton sequence is used to replace the existing rand sequence in the embodiment of the present application, so that the parrot population is densely distributed in the search space, thereby improving the optimization efficiency of the parrot algorithm. The mathematical description of the sequence is:

[0116]

[0117] wherein, H(n) is the initial position of the n-th parrot, n is the ordinal number of the Halton sequence, p is the base of the Halton sequence, is called the root inverse of the initial position n of the parrot with respect to the base p1, and p i (1≤i≤s), wherein s is the number of parameters to be identified, a i is a constant variable.

[0118] Step 402, generating a parrot population;

[0119] Step 403, calculate individual fitness, mark the position of the optimal parrot;

[0120] Step 404, foraging behavior, update the position according to formula (1) and flight strategy formula (2);

[0121] The iteration of the parrot's position in the foraging process is determined by the following formula:

[0122]

[0123] In the formula: is the position vector of the t+1 generation parrot, which represents the updating method of the parrot's position; is the mean value of the position in the t generation parrot population; Levy(d) is a function representing the parrot's flight; best is the position of the current leader parrot.

[0124] The basic parrot algorithm uses Levy flight strategy to calculate the step length in the foraging and resting stages. The large step length generated by Levy flight and the short interval will make the early search insufficient and the late search too tedious, resulting in the lack of precision. At the same time, considering that the length of the step length has a great influence on the accuracy of the algorithm, i.e. large step length helps to explore, and small step length is beneficial to development. Based on this, a tangent flight strategy is adopted in the embodiment of the present application, which uses the tangent to play a gentle role in transition and throughout the global, which helps the parrot algorithm to escape from the local optimal solution. Specifically, the tangent flight strategy is as formula (2).

[0125]

[0126] Where σ represents the parrot flight distance; tan(θ) is the direction of the parrot flight.

[0127] Step 405, resting behavior, update the position according to formula (3) and flight strategy formula (2);

[0128] Resting behavior steps:

[0129]

[0130] Where ones(1, d) represents a vector composed of 1 with dimension d.

[0131] Step 406, communication behavior, update the position according to formula (4);

[0132] Communication behavior includes flying to the bird group and not flying to the bird group. In the parrot algorithm, by introducing a probability control variable p, according to the size of p, and using the mean value of the current parrot population position to represent the center of the group. The mathematical description of this process is:

[0133]

[0134] Step 407, the fear behavior, updates the position according to formula (5);

[0135] The fear behavior: that is, the behavior of the t+1 generation parrot individual subjectively and actively seeking a safe environment due to distrust of the parrot group method. The mathematical description formula of this step is:

[0136]

[0137] Step 408, the parrot individual flies towards the direction of the optimal parrot in the group;

[0138] Step 409, judge whether P is less than 0.1, if yes, execute step 410, if not, execute step 411;

[0139] Step 410, implement the somersault foraging strategy according to formula (6) and formula (7);

[0140] Considering that the parrot algorithm will fly towards the position direction of the optimal parrot in the current group at the end of the settlement, in order to avoid the reduction of the diversity of the parrot flight mode caused thereby, and in order to avoid the parrot algorithm falling into a local extremum, in the embodiment of the application, a somersault action mode based on fuzzy control is used to update the parrot position. The process is described in mathematical language as follows:

[0141] x s (t+1)=x s (t)+S(r1x best -r2x s (t)) (6)

[0142] Wherein, x s represents the t+1 generation parrot position of the s-th parrot; x best is the optimal position in the parrot group; r is a random number (0-1); S is a somersault factor, the value of which is determined by the following formula.

[0143] S=Ks+ε*ΔK s (7)

[0144] Ks is the uncorrected somersault factor, ε represents the proportional factor, and ΔK s represents the somersault factor gain after fuzzy control correction.

[0145] By introducing the fuzzy control mode, the value of the somersault factor can be adaptively changed, that is, in the embodiment of the application, the deviation E of the motor speed ring and its change rate E' are taken as the inputs of the fuzzy controller, and through the design of the fuzzy rule, the fuzzy controller of the somersault factor adaptively performs logical processing, and finally outputs the correction amount ΔK of the somersault factors The fuzzy subsets we selected are {LB, LM, LS, ZE, HS, HM, HB}, which represent a very small, small, small, zero, large, large, very large simulation parameter standard in turn. The determination principle of the fuzzy rules in the embodiment of the present application is: in the initial stage of the motor parameter identification control system, at this time ΔK s is a larger value. As can be seen from formula (7), the larger the air turning factor, the closer the distance between the parrot and the optimal individual, thereby ensuring that the parrot quickly approaches the optimal individual in the initial stage. When the system is in the final stage, at this time the distance between the parrots is already close enough, and the position only needs to be fine-tuned, at this time ΔK s should be a smaller value. In the middle stage, ΔK s may be adjusted according to the response of the system according to actual needs. Finally, the fuzzy rules in the embodiment of the present application are shown in Table 1 as follows:

[0146]

[0147] Table 1

[0148] After being set according to the above provisions, the fuzzy surface generated by the fuzzy rules is shown in Figure 5 . The input variables of the fuzzy surface are the deviation E of the motor speed loop and the rate of change E'. Through the fuzzy surface, it can be directly checked whether the fuzzy rules are set incorrectly. After the fuzzy rules are set, the generated fuzzy surface should be consistent with the one presented in the embodiment of the present application.

[0149] Step 411, updating the individual fitness of the parrot and recording the optimal position at this time; wherein the individual with the smallest fitness is the optimal position;

[0150] Step 412, iteration ends;

[0151] Step 413, outputting the optimal parrot position.

[0152] As can be seen in the embodiment of the present application, as Figure 4 shown in the left dashed frame, the design of the parrot algorithm flow part is presented. When the parrot algorithm logic is introduced in Matlab according to the above description, after the algorithm is entered, the initialization of the parrot algorithm related parameters is first performed, and then the generation step of the parrot population is performed. It should be noted that each parrot is a vector containing the dimensions corresponding to the number of motor parameters to be identified. After the initialization of the parrot population, the algorithm initializes the individual fitness value of the parrot. The individual fitness value is defined as the distance between the parrot individual and the optimal position. After traversing the entire parrot population, the position of the optimal parrot in the population is marked according to the principle of the smallest fitness value.

[0153] Subsequently, the four behaviors of parrot algorithm are performed: 1) foraging behavior, the parrot position is updated according to formula (1) and the improved flight strategy (2); 2) resting behavior, the parrot position is updated according to formula (3) and the improved flight strategy (2); 3) communication behavior, the parrot position is updated according to the principle determined by formula (4); 4) fear behavior, the parrot position is updated according to the principle determined by formula (5). Among them, the parrot individual approaches the direction of the optimal parrot position in the population, that is, the contraction of the solution range of the parrot corresponding to the identified parameter.

[0154] In addition, by introducing the random variable p, the parrot maintains a probability of 0.1 to update the position according to the somersault strategy based on fuzzy control proposed in the application, thereby avoiding the reduction of the diversity of the parrot flight mode at the end of iteration, and at the same time, in order to avoid the parrot algorithm falling into a local extremum; in addition, the parrot individual maintains a probability of 0.9 to calculate the optimal parrot position, thereby updating the size of the individual fitness, so that the position of the optimal parrot in this round of iteration process can be recorded.

[0155] Further, in the parrot interaction motor algorithm process in the embodiment of the application, it is necessary to judge whether the parrot algorithm meets the iteration end condition, in the embodiment of the application, the performance index in the form of formula (8) data is used to decide whether the iteration condition is reached by evaluating the numerical value of the performance index, specifically: taking the integral of the absolute value of the product of time t and the speed loop error E(t) of the permanent magnet synchronous motor system and the sum of t and the current regulation loop error e(t) as the performance index, the function of the performance index is to quantify the good or bad of the motor parameter solved based on the parrot algorithm and returned by calling the motor system model, the mathematical expression of the performance index J is as follows:

[0156]

[0157] In the embodiment of the application, the permanent magnet synchronous motor double closed loop system will continuously generate speed loop error E(t) and current loop error e(t), when the motor system is running, E(t) and e(t) are detected at any time, thereby adaptively generating a series of different sizes of J, the smaller the value is, the smaller the deviation of the speed loop and the current loop of the motor system is, that is, the system reaches steady state at this time, and the identified motor parameters are more accurate. If the parrot algorithm iteration end condition is not met, a new round of iteration of the parrot individual is continued, until the optimal performance index of the system output is reached, at this time, the optimal parrot position vector is the optimal solution set containing the motor to be identified parameters.

[0158] In the embodiment of the present application, the permanent magnet motor control parameter identification strategy is adopted, which adopts the parrot algorithm with simple framework, clear structure and strong algorithm efficiency as the bottom algorithm, and improves the existing parrot algorithm to improve its performance. By setting the performance index as the motor parameter identification strategy for the interaction between the motor parameter identification and the parrot algorithm, the motor parameters identified finally are more accurate.

[0159] Corresponding to the above Figure 1 In the embodiment of the present application, a motor parameter identification device is provided, which is used to execute the following steps:

[0160] S11, initializing the parameters of the parrot algorithm through a preset sequence function, and generating a corresponding parrot population based on the initialized parrot algorithm; wherein the parrot population includes a plurality of parrots, the parrot is a position vector of the motor parameter to be identified, the dimension of the position vector represents the number of the motor parameter to be identified, and the motor parameter to be identified is a motor parameter in a motor double closed loop model;

[0161] S12, determining the individual fitness of each parrot in the parrot population, wherein the individual fitness represents the distance of the parrot from the target position;

[0162] S13, determining the target parrot of the optimal position in the parrot population according to the individual fitness, and updating the position of the parrot according to the target behavior of the parrot, wherein the target behavior represents the behavior of the parrot approaching the target parrot;

[0163] S14, repeatedly executing S12 and S13 until the performance index meets the preset condition to determine the motor parameter corresponding to the position vector, wherein the performance index refers to the performance index of the motor double closed loop model determined based on the parrot with the updated position.

[0164] Based on the device of the embodiment of the present application, the parameters of the parrot algorithm are initialized through a preset sequence function, and the corresponding parrot population is generated based on the initialized parrot algorithm, i.e. the motor parameters to be identified in the motor double closed loop model are mapped to parrots, then the individual fitness of the parrot, i.e. the distance of the parrot from the target position, is determined, then the target parrot of the optimal position in the parrot population is determined through the individual fitness, and the position of the parrot is updated according to the target behavior of the parrot. The updating process is a continuous iteration process, which can continuously update the position of the parrot, improve the identification efficiency of the motor parameter, and improve the identification accuracy of the motor parameter, thereby solving the problems of low accuracy and complex identification process of the motor parameter identification in the related art.

[0165] As Figure 6As shown, the embodiment of the present application provides a device, comprising a processor 611, a communication interface 612, a memory 613 and a communication bus 614, wherein the processor 611, the communication interface 612 and the memory 613 complete mutual communication through the communication bus 614,

[0166] The memory 613 is used for storing a computer program.

[0167] In an embodiment of the present application, the processor 611 is used for executing the program stored in the memory 613, and realizes the motor parameter identification method provided by any one of the foregoing method embodiments, and the roles are similar, which will not be repeated here.

[0168] The embodiment of the present application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize the steps of the motor parameter identification method provided by any one of the foregoing method embodiments.

[0169] The device embodiments described above are only schematic, wherein the units shown as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, that is, they can be located in one place, or distributed on a plurality of network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0170] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be realized by means of software plus a general hardware platform, and of course can also be realized by hardware. Based on such understanding, the above technical solutions can be embodied in the form of software products, and the computer software products can be stored in a computer readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including a plurality of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute the methods described in each embodiment or some parts of the embodiments.

[0171] It is to be understood that the terminology used herein is for the purpose of describing particular example embodiments only and is not intended to be limiting. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. The terms "comprises", "comprising", "includes", "including" and "has" are inclusive and therefore specify the presence of stated features, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, steps, operations, elements, components, and / or groups thereof. The method steps, processes, and operations described herein are not to be construed as necessarily requiring their performance in the particular order

[0172] The above description is merely that of the specific embodiments of the application and as such is not to be taken in a limiting sense. Various modifications and alterations of the embodiments described herein will become apparent to those skilled in the art from the foregoing description, which does not limit the generality presented. It is the intention that all such modifications and alterations be considered equaliy by the spirit and scope of this application. It is therefore intended to cover in the appended claims all such changes and alterations that come within the scope of this application.

Claims

1. A method for identifying parameters of an electrical machine, characterized in that, Comprise: S11, initialize the parameters of parrot algorithm through a preset sequence function, and generate a corresponding parrot population based on the initialized parrot algorithm; wherein the parrot population includes a plurality of parrots, the parrot is a position vector of a to-be-identified motor parameter, the dimension of the position vector represents the number of the to-be-identified motor parameter, and the to-be-identified motor parameter is a motor parameter in a motor double closed loop model; S12, determine the individual fitness of each parrot in the parrot population, wherein the individual fitness represents the distance of the parrot from the target position; S13, determine the target parrot of the optimal position in the parrot population according to the individual fitness, and update the position of the parrot according to the target behavior of the parrot, wherein the target behavior represents the behavior of the parrot approaching the target parrot; S14, repeat S12 and S13 until the performance index meets the preset condition to determine the motor parameter corresponding to the position vector, wherein the performance index refers to the performance index of the motor double closed loop model determined based on the parrot with updated position.

2. The method of claim 1, wherein, Initialize the parameters of parrot algorithm through a preset sequence, comprising: Initialize the parameters of parrot algorithm through Halton sequence; wherein the Halton sequence is ; is the initial position of the n-th parrot, n is the order of the Halton sequence; p is the base of the Halton sequence; and , , is called the root inverse of the parrot initial position n with respect to the base p1, and p i (1≤i≤s), where s is the number of parameters to be identified, a i is a constant variable, denotes the i-th power of the base p1, denotes the inverse of the (i-1)-th power of the base p1. ​ 3. The method of claim 1, wherein, In the case that the target behavior is foraging behavior, updating the position of the parrot according to the target behavior of the parrot, comprising: Determine the flight strategy of the parrot through a first formula, and update the position of the parrot in the foraging process under the flight strategy according to a second formula; The first formula is: wherein, is a position vector of the parrot for the t+1th generation, represents a distance flown by the parrot; is a direction of flight of the parrot; The second formula is: wherein, is the mean of the position within the tth generation of the parrot population, and max is the distance from the target position that is farthest in the parrot population in this iteration of the parrot population; Levy ( d ) a function representing the flight of the parrot; is the position of the lead parrot, and rand represents an arbitrary number in the range [1, ].

4. The method of claim 1, wherein, In the case that the target behavior is staying behavior, updating the position of the parrot according to the target behavior of the parrot, comprising: Determine the flight strategy of the parrot through a first formula, and update the position of the parrot in the staying process under the flight strategy according to a third formula; The first formula is: wherein, is a position vector of the parrot of the t+1th generation, represents a distance flown by the parrot; is a direction of flight of the parrot; The third formula is: wherein represents a set consisting of 1 and having dimension d , is the position of the current lead parrot, Levy d is a function representing the parrot flight situation, is a random function.​ 5. The method of claim 1, wherein, In the case that the target behavior is communication behavior, updating the position of the parrot according to the target behavior of the parrot, comprising: Update the position of the parrot in the communication behavior through a fourth formula; The fourth formula is: wherein, is the probability control variable, is the position vector of the tth generation parrot, is the mean of the positions within the tth generation parrot population, and max is the distance from the target position of the parrot population in this iteration of the parrot population.

6. The method of claim 1, wherein, In the case that the target behavior is fear behavior, updating the position of the parrot according to the target behavior of the parrot, comprising: Update the position of the parrot in the fear through a fifth formula; wherein, is the position vector of the tth generation parrot, is the position of the current leader parrot, and max is the distance of the parrot population from the target position in this round of iteration.

7. The method of claim 1, wherein, Repeat S2 and S3 until the performance index meets the preset condition to determine the motor parameter corresponding to the position vector, comprising: S21, divide the parrots after updating the position obtained by executing S12 and S13 once into parrots of a first preset proportion and parrots of a second preset proportion, wherein the sum of the first preset proportion and the second preset proportion is 1, and the first preset proportion is less than the second preset proportion; S22, update the position of the parrots of the first preset proportion through a sixth formula; The sixth formula is: wherein represents the position of the parrot of the t+1th generation of parrots representing the parrot; is the parrot in the group of parrots that is closest to the target position; and are random numbers between 0 and 1 ; S is an empty translation factor; K s unmodified flip-flop factor, representative scale factor, representative flip-flop factor gain after blur control correction; S23, updating the individual fitness of the first preset proportion of parrots after the position update and the second preset proportion of parrots, and performing S12 again according to the updated individual fitness to re-determine the target parrot; S24, performing the position updating operation of the parrot according to the target behavior of the parrot in S13 again until the performance index meets the preset condition to determine the motor parameter corresponding to the position vector.

8. The method of claim 1, wherein, The performance index is determined in the following manner: determining the integral of the absolute value of the product of the time t and the speed loop error in the motor double closed loop model; taking the sum of the integral and the absolute value of the product of the time and the current regulation loop error as the performance index.

9. An apparatus comprising: at least one communication interface; at least one bus connected with the at least one communication interface; at least one processor connected with the at least one bus; at least one memory connected with the at least one bus, wherein the processor is configured to execute the motor parameter identification method in any one of claims 1 to 8.

10. A computer storage medium storing computer executable instructions for executing the motor parameter identification method in any one of claims 1 to 8.

Images