Surface roughness intelligent measurement system and method applied to motor rotor punching sheet

Abstract

The invention discloses a surface roughness intelligent measurement system and method applied to a motor rotor punching sheet, and relates to the field of roughness measurement. A mapping tool model is provided for subsequent mapping classification of manufacturing parameter data and working state parameter data of the to-be-tested motor rotor punching sheet; the construction of the mapping tool model does not need to directly measure the surface roughness data, so that the subsequent process of optimizing and adjusting the manufacturing parameter data and the working state parameter data of the to-be-measured motor rotor punching sheet is simpler and more convenient, and the operation is more convenient; the manufacturing parameter data and the working state parameter data of the to-be-tested motor rotor punching sheet are adjusted in advance, so that the roughness data obtained through corresponding mapping is maintained at a normal level, the manufacturing process and the working process of the motor rotor punching sheet are optimized, and the service life of the motor rotor punching sheet is longer.

Description

Technical Field

[0001] The invention belongs to the field of roughness measurement, and in particular, relates to an intelligent surface roughness measurement system and method for motor rotor punching sheets. Background Art

[0002] The surface roughness of motor rotor punchings can cause electromagnetic losses in the motor, especially eddy current losses and stray losses; and will increase the discontinuity of magnetic flux and reduce the efficiency of the motor; among them, the surface roughness of motor rotor punchings is directly related to its manufacturing process parameters. If the surface roughness of the manufactured motor rotor punchings cannot be measured in advance and the manufacturing process parameters cannot be reasonably set, the surface roughness of the manufactured motor rotor punchings will be higher, which will affect its application in the motor and cause losses to the company. Summary of the invention

[0003] In view of the problems in the related art, the present invention proposes an intelligent surface roughness measurement system and method for motor rotor punchings to overcome the above technical problems existing in the existing related art.

[0004] In order to solve the above technical problems, the present invention is achieved through the following technical solutions:

[0005] The present invention is an intelligent surface roughness measurement method for motor rotor punchings, comprising the following steps:

[0006] S1. Measuring the roughness data of a plurality of motor rotor punching sheet samples in a stationary state and a working state to obtain an average roughness data set and a second roughness data matrix; collecting manufacturing parameter data and working state parameter data corresponding to a plurality of motor rotor punching sheet samples to obtain a manufacturing parameter data matrix set and a working state parameter data set matrix;

[0007] S2, constructing a final mapping equation according to the manufacturing parameter data matrix set and the average roughness data set; constructing a final SVM classification model according to the working state parameter data set matrix and the second roughness data matrix; collecting the manufacturing parameter data corresponding to the motor rotor punching to be tested, and obtaining the manufacturing parameter data matrix to be tested; collecting the working state parameter data corresponding to the motor rotor punching to be tested, and obtaining the working state parameter data matrix to be tested;

[0008] S3, respectively using the final mapping equation and the final SVM classification model to map and classify the manufacturing parameter data matrix to be measured and the working state parameter data matrix to be measured, to obtain the first roughness data to be measured and the roughness data set to be measured;

[0009] S4, adjusting the manufacturing parameter data matrix to be measured and the working state parameter data matrix to be measured respectively according to the first roughness data to be measured and the roughness data set to be measured;

[0010] In the manufacturing process of motor rotor punchings, the selection of manufacturing parameters has a great influence on the surface roughness of the manufactured motor rotor punchings; at the same time, the state of the motor rotor punchings will also affect its surface roughness. For example, when the motor rotor punchings are in a non-working state, that is, in a stationary state, the surface roughness has a direct impact on the manufacturing parameters set in the manufacturing process; and when the motor rotor punchings are in a working state, the surface roughness is directly related to the working parameters set in its working process; based on this, this scheme first measures the roughness data of multiple motor rotor punching samples in a stationary state and a working state, providing a training basis and a testing basis for the subsequent construction of the final mapping equation and the final SVM classification model; then collects the manufacturing parameter data and the working state parameter data corresponding to the multiple motor rotor punching samples, for the subsequent construction of the manufacturing parameter data and the stationary state. The mapping and classification model between the roughness data and working state parameter data of the motor rotor punching sheet and the roughness data under the working state provides data support; by constructing the final mapping equation and the final SVM classification model, a mapping tool model is provided for the subsequent mapping and classification of the manufacturing parameter data and working state parameter data of the motor rotor punching sheet to be tested; the reason for constructing the mapping tool model is to eliminate the need for direct measurement of the surface roughness data, so that the subsequent optimization and adjustment process of the manufacturing parameter data and working state parameter data of the motor rotor punching sheet to be tested is simpler and easier to operate; by pre-adjusting the manufacturing parameter data and working state parameter data of the motor rotor punching sheet to be tested, the roughness data obtained by the corresponding mapping is maintained at a normal level, thereby optimizing the manufacturing process and working process of the motor rotor punching sheet, and extending the service life of the motor rotor punching sheet.

[0011] Preferably, the S1 comprises the following steps:

[0012] S11, collect multiple motor rotor punching samples to obtain a punching sample set a = {a1, a2, ..., a i ,...,a a′}, a i represents the i-th motor rotor punching sample collected, and a′ represents the total number of motor rotor punching samples collected;

[0013] The punching sample set a={a1, a2, ..., a i ,...,a a′ The roughness data of each motor rotor punching sample in the static state is obtained to obtain the first roughness data matrix Then, a laser interferometer is used to collect the punching sample set a={a1, a2, ..., a i ,...,aa′}, and obtain the second roughness data matrix f; f are as follows:

[0014]

[0015] in, represents the jth roughness data collected from the i-th motor rotor punching sample in a stationary state, represents the total number of roughness data collected for each motor rotor lamination sample; f ij represents the roughness data of the jth historical time point collected for the i-th motor rotor punching sample in the working state, and f′ represents the total number of historical time points corresponding to the roughness data collected for each motor rotor punching sample;

[0016] For the first roughness data matrix The average roughness data of each motor rotor punching sample is calculated to obtain the average roughness data set b = {b1, b2, ..., b i ,...,b a′}, b i Represents the average roughness data of the i-th motor rotor punching sample collected; the calculation formula is as follows,

[0017]

[0018] S12, set the motor rotor punching manufacturing step set b'={b1', b2', b3', b4', b5'} and the working state parameter type set b1′, b2′, b3′, b4′, b5′ respectively represent material cutting step, stamping step, cold working step, heat treatment step and surface treatment step, Indicates the type of the i-th working state parameter set. Indicates the total number of set working state parameter types; sets multiple parameter types in each manufacturing step in the motor rotor punching manufacturing step set b′={b1′, b2′, b3′, b4′, b5′} to obtain a manufacturing parameter type matrix as follows,

[0019]

[0020] in, They represent the jth manufacturing parameter type in the material cutting step, stamping step, cold working step, heat treatment step and surface treatment step, respectively. Indicates the total number of manufacturing parameter types in each motor rotor punching manufacturing step;

[0021] S13, according to the manufacturing parameter type matrix Collect the punching sample set a={a1,a2,...,a i ,...,a a′}, and obtain the manufacturing parameter data matrix set c = {c1, c2, ..., c i ,...,c a′},c i represents the manufacturing parameter data matrix of the i-th motor rotor punching sample collected during the manufacturing process; according to the working state parameter type set Collecting the working state parameter data corresponding to each roughness data in the second roughness data matrix f to obtain the working state parameter data set matrix g; c. i , g are as follows,

[0022]

[0023] Among them, c i1j 、c i2j 、c i3j 、c i4j 、c i5j g respectively represent the manufacturing parameter data corresponding to the jth manufacturing parameter type in the material cutting step, stamping step, cold working step, heat treatment step and surface treatment step of the i-th motor rotor punching sample collected; ij represents f ij The corresponding working status parameter data set, g ijk Indicates g ij The k-th type of working status parameter data;

[0024] Image measurement is a method that uses optical systems and image processing technology to measure the size and shape of objects. It can more accurately measure the surface roughness data of motor rotor punchings in a static state; laser interferometer uses optical or laser technology to calculate the surface roughness by measuring the light reflected or scattered by the surface. It is suitable for fast and non-destructive measurement of surface roughness and can measure the surface roughness data of motor rotor punchings in a working state in real time; the manufacturing parameter type matrix includes punching speed and punching force, etc.; the working state parameter type includes magnetic flux density, current density, iron loss, temperature rise, cooling efficiency, speed, load, voltage, frequency, speed regulation range and torque characteristics, etc.

[0025] Preferably, S2 comprises the following steps:

[0026] S21, according to the manufacturing parameter data matrix set c={c1, c2, ..., c i ,...,c a′}Construct the initial mapping equation; as follows,

[0027]

[0028] Wherein, c′ is the dependent variable of the initial mapping equation, representing the roughness data; is the independent variable of the initial mapping equation, representing the j-th type of manufacturing parameter data of the i-th motor rotor punching manufacturing step; for The independent variable coefficient of , d represents the bias;

[0029] Constructing an initial SVM classification model; using the working state parameter data set matrix g and the second roughness data matrix f to train and test the initial SVM classification model to obtain a final SVM classification model;

[0030] S22, each manufacturing parameter data matrix in the manufacturing parameter data matrix set is substituted into the initial mapping equation to perform mapping calculation, to obtain an initial roughness mapping data set It means that the manufacturing parameter data matrix set c={c1,c2,...,c i ,...,c a′} is substituted into the initial mapping equation to obtain the initial roughness mapping data for mapping calculation;

[0031] Setting a roughness error threshold; when the initial roughness mapping data set has an error between the initial roughness mapping data and the corresponding average roughness data in the average roughness data set that is greater than or equal to the roughness error threshold, optimizing the initial mapping equation until the initial roughness mapping data set does not have an error between the initial roughness mapping data and the corresponding average roughness data in the average roughness data set that is greater than or equal to the roughness error threshold, thereby obtaining a final mapping equation;

[0032] Otherwise, there is no need to optimize the initial mapping equation, and the initial mapping equation is used as the final mapping equation;

[0033] S23, collecting manufacturing parameter data corresponding to the motor rotor punching to be tested according to the manufacturing parameter type matrix, and obtaining the manufacturing parameter data matrix to be tested;

[0034] Collect working state parameter data corresponding to the rotor punching of the motor to be tested according to the working state parameter type set to obtain a data matrix of the working state parameter to be tested;

[0035] This solution uses a multivariate linear equation to construct a mapping relationship between manufacturing parameter data and the corresponding roughness data, making the final mapping equation more concise and easy to calculate, thereby improving the calculation speed when adjusting the manufacturing parameter data corresponding to the motor rotor punching sheet to be tested later; by comparing the data mapped by the initial mapping equation with the data actually measured, the independent variable coefficient and offset of the initial mapping equation are adjusted according to the comparison results, so that the mapping capability of the constructed final mapping equation is better.

[0036] Preferably, in S21, the working state parameter data set matrix g and the second roughness data matrix f are used to train and test the initial SVM classification model, and obtaining the final SVM classification model includes the following steps:

[0037] S211, setting a training data ratio; dividing the working state parameter data set matrix g and the second roughness data matrix f according to the training data ratio to obtain a first working state parameter training data set matrix, a second roughness training data matrix, a first working state parameter test data set matrix, and a second roughness test data matrix;

[0038] S212, setting a training error threshold; inputting the first working state parameter training data set matrix and the second roughness training data matrix into the initial SVM classification model for training; during the training process, when the training error is less than the training error threshold, obtaining a trained SVM classification model; otherwise, continuing the training until the training error is less than the training error threshold;

[0039] S213, setting a test accuracy threshold; inputting the first working state parameter test data set matrix and the second roughness test data matrix into the trained SVM classification model for testing, and obtaining the test accuracy after the test is completed; when the test accuracy is greater than or equal to the test accuracy threshold, the trained SVM classification model is used as the final SVM classification model; otherwise, returning to S212 to continue training the trained SVM classification model until the test accuracy is greater than or equal to the test accuracy threshold;

[0040] By respectively using the first working state parameter training data set matrix, the second roughness training data matrix, the first working state parameter test data set matrix and the second roughness test data matrix to train and test the initial SVM classification model, the final SVM classification model obtained has better classification ability for the working state parameter data.

[0041] Preferably, optimizing the initial mapping equation in S22 comprises the following steps:

[0042] S221. Build the first wild horse population d1′ i represents the i-th wild horse in the first wild horse population, represents the size of the first wild horse population; sets the first leader ratio, and divides the first wild horse population according to the first leader ratio Divide into the first leader wild horse population and the first ordinary wild horse population; set the maximum number of iterations of the first wild horse population to The current number of iterations is are respectively recorded as the first maximum number of iterations and the first current number of iterations; the search space dimension of the first wild horse population is

[0043]

[0044] S222, setting the value range of each independent variable coefficient and the offset in the initial mapping equation, and obtaining the independent variable coefficient value interval matrix e1 and the offset value interval [e 21 ,e 22 ],e 21 、e 22 They represent the upper limit and lower limit of the bias value in the initial mapping equation respectively; e1 is as follows,

[0045]

[0046] in, They respectively represent the lower limits of the values ​​of the independent variables corresponding to the manufacturing parameter data of the j-th manufacturing parameter type in the material cutting step, stamping step, cold working step, heat treatment step and surface treatment step, Respectively represent the upper limits of the values ​​of the independent variables corresponding to the manufacturing parameter data of the j-th manufacturing parameter type in the material cutting step, the stamping step, the cold working step, the heat treatment step, and the surface treatment step;

[0047] According to the independent variable coefficient value interval matrix e1 and the bias value interval [e 21 ,e 22 ] Set the initial position matrix of each wild horse in the first wild horse population, and get the first initial position matrix set e i ' represents the initial position matrix of the i-th wild horse in the first wild horse population; as follows,

[0048]

[0049] Among them, e i ' 1j 、e i ' 2j 、ei ' 3j 、e i ' 4j 、e i ' 5j Respectively represent e i 'The component on the independent variable coefficient dimension in the initial mapping equation corresponding to the manufacturing parameter data corresponding to the j-th manufacturing parameter type in the material cutting step, stamping step, cold working step, heat treatment step and surface treatment step; e' 2i Indicates e i 'The component in the bias dimension in the initial mapping equation;

[0050] S223, according to the average roughness data set b={b1, b2, ..., b i ,...,b a′} and the initial roughness map dataset Constructing the fitness function of the first wild horse population as follows,

[0051]

[0052] Among them, β is a positive number, indicating the correction parameter;

[0053] S224, start iteration, before iteration, set the first current iteration number Set to 1; in the first round of iteration, the first initial position matrix set is used and the fitness function of the first wild horse population is adopted Calculate the fitness value of the initial position matrix of each leader wild horse and ordinary wild horse in the first leader wild horse population and the first ordinary wild horse population to obtain a first fitness value set and a second fitness value set; use the maximum fitness value and the corresponding position matrix in the first fitness value set and the second fitness value set as the first global optimal fitness, the first global optimal position, the second global optimal fitness and the second global optimal position respectively; update the initial position matrix of each leader wild horse and ordinary wild horse in the first leader wild horse population and the first ordinary wild horse population according to the first global optimal fitness, the first global optimal position, the second global optimal fitness and the second global optimal position; after the update is completed, the first current iteration number Add 1 and enter the next iteration;

[0054] In each other iteration, the fitness function of the first wild horse population is used Calculate the fitness value of the position matrix of each leader wild horse and ordinary wild horse in the first leader wild horse population and the first ordinary wild horse population updated in the previous round of iteration to obtain a third fitness value set and a fourth fitness value set; use the maximum fitness value and the corresponding position matrix in the third fitness value set and the fourth fitness value set as the third global optimal fitness, the third global optimal position, the fourth global optimal fitness and the fourth global optimal position respectively; update the position matrix of each leader wild horse and ordinary wild horse in the first leader wild horse population and the first ordinary wild horse population again according to the third global optimal fitness, the third global optimal position, the fourth global optimal fitness and the fourth global optimal position; after the update is completed, the first current iteration number Add 1 and enter the next iteration;

[0055] S225, when When , the iteration is stopped to obtain the first final global optimal position; the first final global optimal position is substituted into the initial mapping equation to obtain the optimized mapping equation; each manufacturing parameter data matrix in the manufacturing parameter data matrix set is substituted into the optimized mapping equation to perform mapping calculation to obtain the optimized roughness mapping data set;

[0056] When the optimized roughness map data set does not have an error between the initial roughness map data and the corresponding average roughness data in the average roughness data set that is greater than or equal to the roughness error threshold, the optimized mapping equation is used as the final mapping equation; otherwise, return to S224 to continue iterating until the optimized roughness map data set does not have an error between the initial roughness map data and the corresponding average roughness data in the average roughness data set that is greater than or equal to the roughness error threshold;

[0057] The Mustang optimization algorithm can effectively explore different areas of the solution space, help to escape from the local optimal solution, find the global optimal solution, and find a better solution within a smaller number of iterations; based on the above, in this scheme, the Mustang optimization algorithm is used to perform multiple iterative optimizations on multiple independent variable coefficients and biases of the initial mapping equation, and the error between the roughness data mapped by the initial mapping equation and the roughness data obtained by actual measurement is used as the fitness function; therefore, as the iteration proceeds, the error between the roughness data mapped by the initial mapping equation and the roughness data obtained by actual measurement becomes smaller and smaller, indicating that the independent variable coefficients and biases obtained by iteration are more accurate.

[0058] Preferably, the S23 comprises the following steps:

[0059] S231, setting the rotor punching of the motor to be tested;

[0060] S232: When the motor rotor punching sheet to be tested is in a stationary state, according to the manufacturing parameter type matrix Collect the manufacturing parameter data of the motor rotor punching to be tested during the manufacturing process to obtain the manufacturing parameter data matrix to be tested as follows,

[0061]

[0062] in, Respectively represent the manufacturing parameter data corresponding to the jth manufacturing parameter type in the material cutting step, stamping step, cold working step, heat treatment step and surface treatment step of the motor rotor punching to be tested;

[0063] The motor rotor punching to be tested is in working state, and working state parameter data of the motor rotor punching to be tested at multiple time points when working are collected according to the working state parameter type set to obtain a working state parameter data matrix to be tested. as follows,

[0064]

[0065] in, represents the data of the working state parameter of the i-th type of the motor rotor punching to be tested at the j-th time point; h represents the total number of historical time points corresponding to the collected working state parameter data of the motor rotor punching to be tested;

[0066] By collecting the manufacturing parameter data and working status parameter data of the motor rotor punching to be tested, a data basis is provided for the subsequent mapping of the surface roughness data of the motor rotor punching to be tested.

[0067] Preferably, S3 comprises the following steps:

[0068] S31, setting a roughness threshold;

[0069] S32, the manufacturing parameter data matrix to be measured Substitute into the final mapping equation for mapping to obtain the first roughness data to be measured Then the working state parameter data matrix to be measured Each column of data is substituted into the final SVM classification model for classification to obtain the roughness data set to be tested;

[0070] By mapping and classifying the manufacturing parameter data and working status parameter data of the motor rotor punching sheet to be tested, the corresponding surface roughness data is obtained, which provides a basis for subsequent determination of whether the manufacturing parameter data and working status parameter data of the motor rotor punching sheet to be tested need to be adjusted.

[0071] Preferably, S4 comprises the following steps:

[0072] S41: when the first roughness data to be measured is greater than or equal to a roughness threshold, Adjustment is performed until the first roughness data to be measured is less than the roughness threshold; otherwise, there is no need to adjust the manufacturing parameter data matrix to be measured. When the roughness data to be measured in the roughness data set to be measured is greater than or equal to the roughness threshold, the working state parameter data matrix to be measured is adjusted. Adjustment is performed until there is no roughness data to be measured greater than or equal to the roughness threshold in the roughness data set to be measured; otherwise, there is no need to adjust the working state parameter data matrix to be measured. Make adjustments;

[0073] Preferably, in S41, the manufacturing parameter data matrix to be measured Making the adjustment involves the following steps:

[0074] S411. Build the second wild horse population d2′ i represents the i-th wild horse in the second wild horse population, Indicates the size of the second wild horse population; sets the second leader ratio, and divides the second wild horse population according to the second leader ratio Divide into the second leader wild horse population and the second ordinary wild horse population; set the maximum number of iterations of the second wild horse population to The current number of iterations is are respectively recorded as the second maximum number of iterations and the second current number of iterations; the search space dimension of the second wild horse population is

[0075] S412: Setting the manufacturing parameter data matrix to be tested The value range of each manufacturing parameter data to be tested is obtained to obtain the parameter data value interval matrix e3; as follows,

[0076]

[0077] in, They respectively represent the lower limits of the manufacturing parameter data corresponding to the j-th manufacturing parameter type in the material cutting step, stamping step, cold working step, heat treatment step and surface treatment step of the motor rotor punching to be tested, Respectively represent the upper limits of the manufacturing parameter data corresponding to the j-th manufacturing parameter type in the material cutting step, stamping step, cold working step, heat treatment step and surface treatment step of the motor rotor punching to be tested;

[0078] The initial position matrix of each wild horse in the second wild horse population is set according to the parameter data value interval matrix e3, and the second initial position matrix set is obtained. h i ' represents the initial position matrix of the i-th wild horse in the second wild horse population; as follows,

[0079]

[0080] Among them, h i ' 1j 、h i ' 2j 、h i ' 3j 、h i ' 4j 、h i ' 5j Respectively represent h i 'The component on the manufacturing parameter data dimension corresponding to the j-th manufacturing parameter type in the material cutting step, stamping step, cold working step, heat treatment step and surface treatment step;

[0081] S413: Based on the first roughness data to be measured Constructing the fitness function of the second wild horse population as follows,

[0082]

[0083] S414, start iteration, before iteration, set the second current iteration number Set to 1; in the first round of iteration, the second initial position matrix set and the final mapping equation are used and the fitness function of the second wild horse population is adopted Calculate the fitness value of the initial position matrix of each leader wild horse and ordinary wild horse in the second leader wild horse population and the second ordinary wild horse population to obtain the fifth fitness value set and the sixth fitness value set; use the maximum fitness value and the corresponding position matrix in the fifth fitness value set and the sixth fitness value set as the fifth global optimal fitness, the fifth global optimal position, the sixth global optimal fitness and the sixth global optimal position respectively; update the initial position matrix of each leader wild horse and ordinary wild horse in the second leader wild horse population and the second ordinary wild horse population according to the fifth global optimal fitness, the fifth global optimal position, the sixth global optimal fitness and the sixth global optimal position; after the update is completed, the second current iteration number Add 1 and enter the next iteration;

[0084] In each other round of iteration, the final mapping equation is matched and the fitness function of the second wild horse population is adopted Calculate the fitness value of the position matrix of each leader wild horse and ordinary wild horse in the second leader wild horse population and the second ordinary wild horse population updated in the previous round of iteration to obtain the seventh fitness value set and the eighth fitness value set; use the maximum fitness value and the corresponding position matrix in the seventh fitness value set and the eighth fitness value set as the seventh global optimal fitness, the seventh global optimal position, the eighth global optimal fitness and the eighth global optimal position respectively; update the position matrix of each leader wild horse and ordinary wild horse in the second leader wild horse population and the second ordinary wild horse population again according to the seventh global optimal fitness, the seventh global optimal position, the eighth global optimal fitness and the eighth global optimal position; after the update is completed, the second current iteration number Add 1 and enter the next iteration;

[0085] S415, when When , the iteration is stopped to obtain the second final global optimal position; the second final global optimal position is used as the adjusted first roughness data to be measured; when the adjusted first roughness data to be measured is less than the roughness threshold, the adjustment is completed; otherwise, return to S414 to continue iterating until the adjusted first roughness data to be measured is less than the roughness threshold;

[0086] The Mustang optimization algorithm is used to perform multiple iterative optimizations on the manufacturing parameter data of the motor rotor punching to be tested, and the size of the surface roughness data obtained by the iteration is used as the fitness function; therefore, as the iteration proceeds, the surface roughness data of the motor rotor punching to be tested obtained by mapping becomes smaller and smaller, indicating that the manufacturing parameter data obtained by the iteration is better.

[0087] The surface roughness intelligent measurement system applied to motor rotor punching sheets includes a roughness data measurement module, a motor rotor punching sheet manufacturing parameter type setting module, a first manufacturing parameter data acquisition module, a first working state parameter data acquisition module, a mapping equation construction module, a classification model construction module, a second manufacturing parameter data acquisition module, a second working state parameter data acquisition module, a mapping classification model and a parameter data adjustment module;

[0088] The roughness data measurement module is used to measure the roughness data of multiple motor rotor punching samples in a static state and a working state to obtain an average roughness data set and a second roughness data matrix;

[0089] The motor rotor punching sheet manufacturing parameter type setting module is used to set the manufacturing parameter types and working state parameter types of multiple motor rotor punching sheet manufacturing steps to obtain a manufacturing parameter type matrix and a working state parameter type set;

[0090] The first manufacturing parameter data acquisition module is used to acquire manufacturing parameter data corresponding to a plurality of motor rotor punching samples according to a manufacturing parameter type matrix to obtain a manufacturing parameter data matrix set;

[0091] The first working state parameter data acquisition module is used to acquire working state parameter data corresponding to a plurality of motor rotor punching samples according to a working state parameter type set to obtain a working state parameter data set matrix;

[0092] The mapping equation building module is used to build a final mapping equation according to the manufacturing parameter data matrix set and the average roughness data set;

[0093] The classification model building module is used to build a final SVM classification model according to the working state parameter data set matrix and the second roughness data matrix;

[0094] The second manufacturing parameter data acquisition module is used to acquire the manufacturing parameter data corresponding to the motor rotor punching to be tested according to the manufacturing parameter type matrix to obtain the manufacturing parameter data matrix to be tested;

[0095] The second working state parameter data acquisition module is used to acquire working state parameter data corresponding to the motor rotor punching to be tested according to the working state parameter type set to obtain a working state parameter data matrix to be tested;

[0096] The mapping classification model is used to respectively use the final mapping equation and the final SVM classification model to map and classify the manufacturing parameter data matrix to be measured and the working state parameter data matrix to be measured, so as to obtain the first roughness data to be measured and the roughness data set to be measured;

[0097] The parameter data adjustment module is used to adjust the manufacturing parameter data matrix to be measured and the working state parameter data matrix to be measured according to the first roughness data to be measured and the roughness data set to be measured.

[0098] The present invention has the following beneficial effects:

[0099] 1. In the present invention, by constructing a final mapping equation and a final SVM classification model, a mapping tool model is provided for subsequent mapping and classification of the manufacturing parameter data and the working state parameter data of the motor rotor punching to be tested; the reason for constructing the mapping tool model is to eliminate the need for direct measurement of the surface roughness data, so that the subsequent process of optimizing and adjusting the manufacturing parameter data and the working state parameter data of the motor rotor punching to be tested is simpler and more convenient to operate; by pre-adjusting the manufacturing parameter data and the working state parameter data of the motor rotor punching to be tested, the roughness data obtained by the corresponding mapping is maintained at a normal level, thereby optimizing the manufacturing process and the working process of the motor rotor punching, and making the service life of the motor rotor punching longer.

[0100] 2. In the present invention, the data mapped by the initial mapping equation is compared with the data actually measured, and the independent variable coefficient and the bias of the initial mapping equation are adjusted according to the comparison result, so that the mapping ability of the final mapping equation constructed is better.

[0101] 3. In the present invention, the Mustang optimization algorithm is used to perform multiple iterative optimizations on the manufacturing parameter data of the motor rotor punching to be tested, and the size of the surface roughness data obtained by the iteration is used as the fitness function; therefore, as the iteration proceeds, the surface roughness data of the motor rotor punching to be tested obtained by mapping becomes smaller and smaller, indicating that the manufacturing parameter data obtained by the iteration is better.

[0102] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0103] In order to more clearly illustrate the technical solutions of the embodiments of the invention, the drawings required for describing the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the invention. For ordinary technicians in this field, they can also obtain drawings based on these drawings without paying creative work.

[0104] Figure 1 The figure is a flow chart of the intelligent measurement method of the surface roughness of motor rotor punching sheets applied to the present invention. DETAILED DESCRIPTION

[0105] The following will be combined with the drawings in the embodiments of the invention to clearly and completely describe the technical solutions in the embodiments of the invention. Obviously, the described embodiments are only part of the embodiments of the invention, not all of the embodiments. Based on the embodiments in the invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the invention.

[0106] In the description of the present invention, it is necessary to understand that the terms "opening", "upper", "lower", "top", "middle", "inside" and the like indicating orientation or positional relationship are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the components or elements referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as limiting the invention.

[0107] Embodiment 1

[0108] This embodiment is an intelligent surface roughness measurement method applied to motor rotor punchings, comprising the following steps:

[0109] S1. Measuring the roughness data of a plurality of motor rotor punching sheet samples in a stationary state and a working state to obtain an average roughness data set and a second roughness data matrix; collecting manufacturing parameter data and working state parameter data corresponding to a plurality of motor rotor punching sheet samples to obtain a manufacturing parameter data matrix set and a working state parameter data set matrix;

[0110] The S1 comprises the following steps:

[0111] S11, collect multiple motor rotor punching samples to obtain a punching sample set a = {a1, a2, ..., a i ,...,a a′}, a i represents the i-th motor rotor punching sample collected, and a′ represents the total number of motor rotor punching samples collected;

[0112] The punching sample set a={a1, a2, ..., a i ,...,a a′ The roughness data of each motor rotor punching sample in the static state is obtained to obtain the first roughness data matrix Then, a laser interferometer is used to collect the punching sample set a={a1, a2, ..., a i ,...,a a′}, and obtain the second roughness data matrix f; f are as follows:

[0113]

[0114] in, represents the jth roughness data collected from the i-th motor rotor punching sample in a stationary state, represents the total number of roughness data collected for each motor rotor lamination sample; f ij represents the roughness data of the jth historical time point collected for the i-th motor rotor punching sample in the working state, and f′ represents the total number of historical time points corresponding to the roughness data collected for each motor rotor punching sample;

[0115] For the first roughness data matrix The average roughness data of each motor rotor punching sample is calculated to obtain the average roughness data set b = {b1, b2, ..., b i ,...,b a′}, b i Represents the average roughness data of the i-th motor rotor punching sample collected; the calculation formula is as follows,

[0116]

[0117] S12, set the motor rotor punching manufacturing step set b'={b1', b2', b3', b4', b5'} and the working state parameter type set b1′, b2′, b3′, b4′, b5′ respectively represent material cutting step, stamping step, cold working step, heat treatment step and surface treatment step, Indicates the type of the i-th working state parameter set. Indicates the total number of set working state parameter types; sets multiple parameter types in each manufacturing step in the motor rotor punching manufacturing step set b′={b1′, b2′, b3′, b4′, b5′} to obtain a manufacturing parameter type matrix as follows,

[0118]

[0119] in, They represent the jth manufacturing parameter type in the material cutting step, stamping step, cold working step, heat treatment step and surface treatment step, respectively. Indicates the total number of manufacturing parameter types in each motor rotor punching manufacturing step;

[0120] S13, according to the manufacturing parameter type matrix Collect the punching sample set a={a1,a2,...,a i ,...,a a′}, and obtain the manufacturing parameter data matrix set c = {c1, c2, ..., c i ,...,c a′},c i represents the manufacturing parameter data matrix of the i-th motor rotor punching sample collected during the manufacturing process; according to the working state parameter type set Collecting the working state parameter data corresponding to each roughness data in the second roughness data matrix f to obtain the working state parameter data set matrix g; c. i , g are as follows,

[0121]

[0122] Among them, c i1j 、c i2j 、c i3j 、c i4j 、c i5jg respectively represent the manufacturing parameter data corresponding to the jth manufacturing parameter type in the material cutting step, stamping step, cold working step, heat treatment step and surface treatment step of the i-th motor rotor punching sample collected; ij represents f ij The corresponding working status parameter data set,

[0123] g ijk Indicates g ij The k-th type of working status parameter data;

[0124] S2, constructing a final mapping equation according to the manufacturing parameter data matrix set and the average roughness data set; constructing a final SVM classification model according to the working state parameter data set matrix and the second roughness data matrix; collecting the manufacturing parameter data corresponding to the motor rotor punching to be tested, and obtaining the manufacturing parameter data matrix to be tested; collecting the working state parameter data corresponding to the motor rotor punching to be tested, and obtaining the working state parameter data matrix to be tested;

[0125] The S2 comprises the following steps:

[0126] S21, according to the manufacturing parameter data matrix set c={c1, c2, ..., c i ,...,c a′}Construct the initial mapping equation; as follows,

[0127]

[0128] Wherein, c′ is the dependent variable of the initial mapping equation, representing the roughness data; is the independent variable of the initial mapping equation, representing the j-th type of manufacturing parameter data of the i-th motor rotor punching manufacturing step; for The independent variable coefficient of , d represents the bias;

[0129] Constructing an initial SVM classification model; using the working state parameter data set matrix g and the second roughness data matrix f to train and test the initial SVM classification model to obtain a final SVM classification model;

[0130] In S21, the initial SVM classification model is trained and tested using the working state parameter data set matrix g and the second roughness data matrix f, and obtaining the final SVM classification model includes the following steps:

[0131] S211, setting a training data ratio; dividing the working state parameter data set matrix g and the second roughness data matrix f according to the training data ratio to obtain a first working state parameter training data set matrix, a second roughness training data matrix, a first working state parameter test data set matrix, and a second roughness test data matrix;

[0132] S212, setting a training error threshold; inputting the first working state parameter training data set matrix and the second roughness training data matrix into the initial SVM classification model for training; during the training process, when the training error is less than the training error threshold, obtaining a trained SVM classification model; otherwise, continuing the training until the training error is less than the training error threshold;

[0133] S213, setting a test accuracy threshold; inputting the first working state parameter test data set matrix and the second roughness test data matrix into the trained SVM classification model for testing, and obtaining the test accuracy after the test is completed; when the test accuracy is greater than or equal to the test accuracy threshold, the trained SVM classification model is used as the final SVM classification model; otherwise, returning to S212 to continue training the trained SVM classification model until the test accuracy is greater than or equal to the test accuracy threshold;

[0134] S22, each manufacturing parameter data matrix in the manufacturing parameter data matrix set is substituted into the initial mapping equation to perform mapping calculation, to obtain an initial roughness mapping data set It means that the manufacturing parameter data matrix set c={c1,c2,...,c i ,...,c a′} is substituted into the initial mapping equation to obtain the initial roughness mapping data for mapping calculation;

[0135] Setting a roughness error threshold; when the initial roughness mapping data set has an error between the initial roughness mapping data and the corresponding average roughness data in the average roughness data set that is greater than or equal to the roughness error threshold, optimizing the initial mapping equation until the initial roughness mapping data set does not have an error between the initial roughness mapping data and the corresponding average roughness data in the average roughness data set that is greater than or equal to the roughness error threshold, thereby obtaining a final mapping equation;

[0136] Otherwise, there is no need to optimize the initial mapping equation, and the initial mapping equation is used as the final mapping equation;

[0137] Optimizing the initial mapping equation in S22 includes the following steps:

[0138] S221. Build the first wild horse population d1′ i represents the i-th wild horse in the first wild horse population, represents the size of the first wild horse population; sets the first leader ratio, and divides the first wild horse population according to the first leader ratio Divide into the first leader wild horse population and the first ordinary wild horse population; set the maximum number of iterations of the first wild horse population to The current number of iterations is are respectively recorded as the first maximum number of iterations and the first current number of iterations; the search space dimension of the first wild horse population is

[0139]

[0140] S222, setting the value range of each independent variable coefficient and the offset in the initial mapping equation, and obtaining the independent variable coefficient value interval matrix e1 and the offset value interval [e 21 ,e 22 ],e 21 、e 22 They represent the upper limit and lower limit of the bias value in the initial mapping equation respectively; e1 is as follows,

[0141]

[0142] in, They respectively represent the lower limits of the values ​​of the independent variables corresponding to the manufacturing parameter data of the j-th manufacturing parameter type in the material cutting step, stamping step, cold working step, heat treatment step and surface treatment step, Respectively represent the upper limits of the values ​​of the independent variables corresponding to the manufacturing parameter data of the j-th manufacturing parameter type in the material cutting step, the stamping step, the cold working step, the heat treatment step, and the surface treatment step;

[0143] According to the independent variable coefficient value interval matrix e1 and the bias value interval [e 21 ,e 22 ] Set the initial position matrix of each wild horse in the first wild horse population, and get the first initial position matrix set e i ' represents the initial position matrix of the i-th wild horse in the first wild horse population; as follows,

[0144]

[0145] Among them, e i ' 1j、e i ' 2j 、e i ' 3j 、e i ' 4j 、e i ' 5j Respectively represent e i 'The component on the independent variable coefficient dimension in the initial mapping equation corresponding to the manufacturing parameter data corresponding to the j-th manufacturing parameter type in the material cutting step, stamping step, cold working step, heat treatment step and surface treatment step; e' 2i Indicates e i 'The component in the bias dimension in the initial mapping equation;

[0146] S223, according to the average roughness data set b={b1, b2, ..., b i ,...,b a′} and the initial roughness map dataset Constructing the fitness function of the first wild horse population as follows,

[0147]

[0148] Among them, β is a positive number, indicating the correction parameter;

[0149] S224, start iteration, before iteration, set the first current iteration number Set to 1; in the first round of iteration, the first initial position matrix set is used and the fitness function of the first wild horse population is adopted Calculate the fitness value of the initial position matrix of each leader wild horse and ordinary wild horse in the first leader wild horse population and the first ordinary wild horse population to obtain a first fitness value set and a second fitness value set; use the maximum fitness value and the corresponding position matrix in the first fitness value set and the second fitness value set as the first global optimal fitness, the first global optimal position, the second global optimal fitness and the second global optimal position respectively; update the initial position matrix of each leader wild horse and ordinary wild horse in the first leader wild horse population and the first ordinary wild horse population according to the first global optimal fitness, the first global optimal position, the second global optimal fitness and the second global optimal position; after the update is completed, the first current iteration number Add 1 and enter the next iteration;

[0150] In each other iteration, the fitness function of the first wild horse population is used Calculate the fitness value of the position matrix of each leader wild horse and ordinary wild horse in the first leader wild horse population and the first ordinary wild horse population updated in the previous round of iteration to obtain a third fitness value set and a fourth fitness value set; use the maximum fitness value and the corresponding position matrix in the third fitness value set and the fourth fitness value set as the third global optimal fitness, the third global optimal position, the fourth global optimal fitness and the fourth global optimal position respectively; update the position matrix of each leader wild horse and ordinary wild horse in the first leader wild horse population and the first ordinary wild horse population again according to the third global optimal fitness, the third global optimal position, the fourth global optimal fitness and the fourth global optimal position; after the update is completed, the first current iteration number Add 1 and enter the next iteration;

[0151] S225, when When , the iteration is stopped to obtain the first final global optimal position; the first final global optimal position is substituted into the initial mapping equation to obtain the optimized mapping equation; each manufacturing parameter data matrix in the manufacturing parameter data matrix set is substituted into the optimized mapping equation to perform mapping calculation to obtain the optimized roughness mapping data set;

[0152] When the optimized roughness map data set does not have an error between the initial roughness map data and the corresponding average roughness data in the average roughness data set that is greater than or equal to the roughness error threshold, the optimized mapping equation is used as the final mapping equation; otherwise, return to S224 to continue iterating until the optimized roughness map data set does not have an error between the initial roughness map data and the corresponding average roughness data in the average roughness data set that is greater than or equal to the roughness error threshold;

[0153] S23, collecting manufacturing parameter data corresponding to the motor rotor punching to be tested according to the manufacturing parameter type matrix, and obtaining the manufacturing parameter data matrix to be tested;

[0154] Collect working state parameter data corresponding to the rotor punching of the motor to be tested according to the working state parameter type set to obtain a data matrix of the working state parameter to be tested;

[0155] The S23 comprises the following steps:

[0156] S231, setting the rotor punching of the motor to be tested;

[0157] S232: When the motor rotor punching sheet to be tested is in a stationary state, according to the manufacturing parameter type matrix Collect the manufacturing parameter data of the motor rotor punching to be tested during the manufacturing process to obtain the manufacturing parameter data matrix to be tested as follows,

[0158]

[0159] in, Respectively represent the manufacturing parameter data corresponding to the jth manufacturing parameter type in the material cutting step, stamping step, cold working step, heat treatment step and surface treatment step of the motor rotor punching to be tested;

[0160] The motor rotor punching to be tested is in working state, and working state parameter data of the motor rotor punching to be tested at multiple time points when working are collected according to the working state parameter type set to obtain a working state parameter data matrix to be tested. as follows,

[0161]

[0162] in, represents the data of the working state parameter of the i-th type of the motor rotor punching to be tested at the j-th time point; h represents the total number of historical time points corresponding to the collected working state parameter data of the motor rotor punching to be tested;

[0163] S3, respectively using the final mapping equation and the final SVM classification model to map and classify the manufacturing parameter data matrix to be measured and the working state parameter data matrix to be measured, to obtain the first roughness data to be measured and the roughness data set to be measured;

[0164] The S3 comprises the following steps:

[0165] S31, setting a roughness threshold;

[0166] S32, the manufacturing parameter data matrix to be measured Substitute into the final mapping equation for mapping to obtain the first roughness data to be measured Then the working state parameter data matrix to be measured Each column of data is substituted into the final SVM classification model for classification to obtain the roughness data set to be tested;

[0167] S4, adjusting the manufacturing parameter data matrix to be measured and the working state parameter data matrix to be measured respectively according to the first roughness data to be measured and the roughness data set to be measured;

[0168] S41: when the first roughness data to be measured is greater than or equal to a roughness threshold, Adjustment is performed until the first roughness data to be measured is less than the roughness threshold; otherwise, there is no need to adjust the manufacturing parameter data matrix to be measured. When the roughness data to be measured in the roughness data set to be measured is greater than or equal to the roughness threshold, the working state parameter data matrix to be measured is adjusted. Adjustment is performed until there is no roughness data to be measured greater than or equal to the roughness threshold in the roughness data set to be measured; otherwise, there is no need to adjust the working state parameter data matrix to be measured. Make adjustments;

[0169] In S41, the manufacturing parameter data matrix to be measured is Making the adjustment involves the following steps:

[0170] S411. Build the second wild horse population d2′ i represents the i-th wild horse in the second wild horse population, Indicates the size of the second wild horse population; sets the second leader ratio, and divides the second wild horse population according to the second leader ratio Divide into the second leader wild horse population and the second ordinary wild horse population; set the maximum number of iterations of the second wild horse population to The current number of iterations is are respectively recorded as the second maximum number of iterations and the second current number of iterations; the search space dimension of the second wild horse population is

[0171] S412: Setting the manufacturing parameter data matrix to be tested The value range of each manufacturing parameter data to be tested is obtained to obtain the parameter data value interval matrix e3; as follows,

[0172]

[0173] in, They respectively represent the lower limits of the manufacturing parameter data corresponding to the j-th manufacturing parameter type in the material cutting step, stamping step, cold working step, heat treatment step and surface treatment step of the motor rotor punching to be tested, Respectively represent the upper limits of the manufacturing parameter data corresponding to the j-th manufacturing parameter type in the material cutting step, stamping step, cold working step, heat treatment step and surface treatment step of the motor rotor punching to be tested;

[0174] The initial position matrix of each wild horse in the second wild horse population is set according to the parameter data value interval matrix e3, and the second initial position matrix set is obtained. h i ' represents the initial position matrix of the i-th wild horse in the second wild horse population; as follows,

[0175]

[0176] Among them, h i ' 1j 、h i ' 2j 、h i ' 3j 、h i ' 4j 、h i ' 5j Respectively represent h i 'The component on the manufacturing parameter data dimension corresponding to the j-th manufacturing parameter type in the material cutting step, stamping step, cold working step, heat treatment step and surface treatment step;

[0177] S413: Based on the first roughness data to be measured Constructing the fitness function of the second wild horse population as follows,

[0178]

[0179] S414, start iteration, before iteration, set the second current iteration number Set to 1; in the first round of iteration, the second initial position matrix set and the final mapping equation are used and the fitness function of the second wild horse population is adopted Calculate the fitness value of the initial position matrix of each leader wild horse and ordinary wild horse in the second leader wild horse population and the second ordinary wild horse population to obtain the fifth fitness value set and the sixth fitness value set; use the maximum fitness value and the corresponding position matrix in the fifth fitness value set and the sixth fitness value set as the fifth global optimal fitness, the fifth global optimal position, the sixth global optimal fitness and the sixth global optimal position respectively; update the initial position matrix of each leader wild horse and ordinary wild horse in the second leader wild horse population and the second ordinary wild horse population according to the fifth global optimal fitness, the fifth global optimal position, the sixth global optimal fitness and the sixth global optimal position; after the update is completed, the second current iteration number Add 1 and enter the next iteration;

[0180] In each other round of iteration, the final mapping equation is matched and the fitness function of the second wild horse population is adopted Calculate the fitness value of the position matrix of each leader wild horse and ordinary wild horse in the second leader wild horse population and the second ordinary wild horse population updated in the previous round of iteration to obtain the seventh fitness value set and the eighth fitness value set; use the maximum fitness value and the corresponding position matrix in the seventh fitness value set and the eighth fitness value set as the seventh global optimal fitness, the seventh global optimal position, the eighth global optimal fitness and the eighth global optimal position respectively; update the position matrix of each leader wild horse and ordinary wild horse in the second leader wild horse population and the second ordinary wild horse population again according to the seventh global optimal fitness, the seventh global optimal position, the eighth global optimal fitness and the eighth global optimal position; after the update is completed, the second current iteration number Add 1 and enter the next iteration;

[0181] S415, when , stop iteration and obtain the second final global optimal position; use the second final global optimal position as the adjusted first roughness data to be measured; when the adjusted first roughness data to be measured is less than the roughness threshold, the adjustment is completed; otherwise, return to S414 and continue to iterate until the adjusted first roughness data to be measured is less than the roughness threshold.

[0182] Embodiment 2

[0183] This embodiment discloses an intelligent surface roughness measurement system for motor rotor punching sheets, and the system can implement the method of the above embodiment, including a roughness data measurement module, a motor rotor punching sheet manufacturing parameter type setting module, a first manufacturing parameter data acquisition module, a first working state parameter data acquisition module, a mapping equation construction module, a classification model construction module, a second manufacturing parameter data acquisition module, a second working state parameter data acquisition module, a mapping classification model and a parameter data adjustment module;

[0184] The roughness data measurement module is used to measure the roughness data of multiple motor rotor punching samples in a static state and a working state to obtain an average roughness data set and a second roughness data matrix;

[0185] The motor rotor punching sheet manufacturing parameter type setting module is used to set the manufacturing parameter types and working state parameter types of multiple motor rotor punching sheet manufacturing steps to obtain a manufacturing parameter type matrix and a working state parameter type set;

[0186] The first manufacturing parameter data acquisition module is used to acquire manufacturing parameter data corresponding to a plurality of motor rotor punching samples according to a manufacturing parameter type matrix to obtain a manufacturing parameter data matrix set;

[0187] The first working state parameter data acquisition module is used to acquire working state parameter data corresponding to a plurality of motor rotor punching samples according to a working state parameter type set to obtain a working state parameter data set matrix;

[0188] The mapping equation building module is used to build a final mapping equation according to the manufacturing parameter data matrix set and the average roughness data set;

[0189] The classification model building module is used to build a final SVM classification model according to the working state parameter data set matrix and the second roughness data matrix;

[0190] The second manufacturing parameter data acquisition module is used to acquire the manufacturing parameter data corresponding to the motor rotor punching to be tested according to the manufacturing parameter type matrix to obtain the manufacturing parameter data matrix to be tested;

[0191] The second working state parameter data acquisition module is used to acquire working state parameter data corresponding to the motor rotor punching to be tested according to the working state parameter type set to obtain a working state parameter data matrix to be tested;

[0192] The mapping classification model is used to respectively use the final mapping equation and the final SVM classification model to map and classify the manufacturing parameter data matrix to be measured and the working state parameter data matrix to be measured, so as to obtain the first roughness data to be measured and the roughness data set to be measured;

[0193] The parameter data adjustment module is used to adjust the manufacturing parameter data matrix to be measured and the working state parameter data matrix to be measured according to the first roughness data to be measured and the roughness data set to be measured.

[0194] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0195] The preferred embodiments of the invention disclosed above are only used to help explain the invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the invention, so that those skilled in the art can understand and use the invention well.

Claims

1. An intelligent surface roughness measurement method for motor rotor punchings, characterized in that: The following steps are involved: S1. Measuring the roughness data of a plurality of motor rotor punching sheet samples in a stationary state and a working state to obtain an average roughness data set and a second roughness data matrix; collecting manufacturing parameter data and working state parameter data corresponding to a plurality of motor rotor punching sheet samples to obtain a manufacturing parameter data matrix set and a working state parameter data set matrix; S2, constructing the final mapping equation according to the manufacturing parameter data matrix set and the average roughness data set; Construct the final SVM classification model according to the working state parameter data set matrix and the second roughness data matrix; Collecting manufacturing parameter data corresponding to the motor rotor punching to be tested, and obtaining a manufacturing parameter data matrix to be tested; collecting working state parameter data corresponding to the motor rotor punching to be tested, and obtaining a working state parameter data matrix to be tested; S3, respectively using the final mapping equation and the final SVM classification model to map and classify the manufacturing parameter data matrix to be measured and the working state parameter data matrix to be measured, to obtain the first roughness data to be measured and the roughness data set to be measured; S4. Adjust the manufacturing parameter data matrix to be measured and the working state parameter data matrix to be measured respectively according to the first roughness data to be measured and the roughness data set to be measured.

2. The surface roughness intelligent measurement method for motor rotor punching according to claim 1 is characterized in that: The S1 comprises the following steps: S11, collecting multiple motor rotor punching samples to obtain a punching sample set; using an image measurement method to collect roughness data of each motor rotor punching sample in a static state for multiple times to obtain a first roughness data matrix; then using a laser interferometer to collect roughness data of each motor rotor punching sample in the punching sample set in a working state at multiple historical time points to obtain a second roughness data matrix; calculating the average value of multiple roughness data of each motor rotor punching sample in the first roughness data matrix to obtain an average roughness data set; S12, setting a motor rotor punching manufacturing step set and a working state parameter type set; setting multiple parameter types in each manufacturing step in the motor rotor punching manufacturing step set to obtain a manufacturing parameter type matrix; S13. According to the manufacturing parameter type matrix, the manufacturing parameter data of each motor rotor punching sample in the punching sample set during the manufacturing process are collected to obtain a manufacturing parameter data matrix set; according to the working state parameter type set, the working state parameter data corresponding to each roughness data in the second roughness data matrix are collected to obtain a working state parameter data set matrix.

3. The intelligent surface roughness measurement method for motor rotor punching according to claim 2 is characterized in that: The S2 comprises the following steps: S21, constructing an initial mapping equation according to the manufacturing parameter data matrix set; Constructing an initial SVM classification model; using the working state parameter data set matrix and the second roughness data matrix to train and test the initial SVM classification model to obtain a final SVM classification model; S22, respectively substituting each manufacturing parameter data matrix in the manufacturing parameter data matrix set into the initial mapping equation for mapping calculation to obtain an initial roughness mapping data set; setting a roughness error threshold; when there is an error between the initial roughness mapping data in the initial roughness mapping data set and the corresponding average roughness data in the average roughness data set that is greater than or equal to the roughness error threshold, optimizing the initial mapping equation until there is no error between the initial roughness mapping data in the initial roughness mapping data set and the corresponding average roughness data in the average roughness data set that is greater than or equal to the roughness error threshold, thereby obtaining a final mapping equation; otherwise, there is no need to optimize the initial mapping equation, and the initial mapping equation is used as the final mapping equation; S23, collecting manufacturing parameter data corresponding to the motor rotor punching to be tested according to the manufacturing parameter type matrix to obtain the manufacturing parameter data matrix to be tested; collecting working state parameter data corresponding to the motor rotor punching to be tested according to the working state parameter type set to obtain the working state parameter data matrix to be tested.

4. The intelligent surface roughness measurement method for motor rotor punching according to claim 3 is characterized in that: In S22, the initial mapping equation is optimized using the Mustang optimization algorithm.

5. The intelligent surface roughness measurement method for motor rotor punching according to claim 4 is characterized in that: The S23 comprises the following steps: S231, setting the rotor punching of the motor to be tested; S232. When the motor rotor punching to be tested is in a stationary state, the manufacturing parameter data of the motor rotor punching to be tested during the manufacturing process are collected according to the manufacturing parameter type matrix to obtain the manufacturing parameter data matrix to be tested; and the working state parameter data of the motor rotor punching to be tested at multiple time points during operation are collected according to the working state parameter type set to obtain the working state parameter data matrix to be tested.

6. The intelligent surface roughness measurement method for motor rotor punching according to claim 5 is characterized in that: The S3 comprises the following steps: S31, setting a roughness threshold; S32, substituting the manufacturing parameter data matrix to be measured into the final mapping equation for mapping, and obtaining the first roughness data to be measured; and then substituting each column of the working state parameter data matrix to be measured into the final SVM classification model for classification, and obtaining the roughness data set to be measured.

7. The intelligent surface roughness measurement method for motor rotor punchings according to claim 6 is characterized in that: The S4 comprises the following steps: S41. When the first roughness data to be measured is greater than or equal to the roughness threshold, the manufacturing parameter data matrix to be measured is adjusted until the first roughness data to be measured is less than the roughness threshold; otherwise, the manufacturing parameter data matrix to be measured does not need to be adjusted; when there is roughness data to be measured greater than or equal to the roughness threshold in the roughness data set to be measured, the working state parameter data matrix to be measured is adjusted until there is no roughness data to be measured greater than or equal to the roughness threshold in the roughness data set to be measured; otherwise, the working state parameter data matrix to be measured does not need to be adjusted.

8. The intelligent surface roughness measurement method for motor rotor punchings according to claim 7 is characterized in that: The step of adjusting the manufacturing parameter data matrix to be measured in S41 includes the following steps: S411, construct a second wild horse population; set a second leader ratio, and divide the second wild horse population into a second leader wild horse population and a second common wild horse population according to the second leader ratio; set the maximum number of iterations of the second wild horse population to The current number of iterations is are recorded as the second maximum number of iterations and the second current number of iterations respectively; S412, setting the value range of each manufacturing parameter data to be tested in the manufacturing parameter data matrix to obtain a parameter data value interval matrix; setting the initial position matrix of each wild horse in the second wild horse population according to the parameter data value interval matrix to obtain a second initial position matrix set; S413, constructing a fitness function of a second wild horse population according to the first roughness data to be measured; S414, start iteration; in each round of iteration, the fitness value of the position matrix of each leader wild horse and common wild horses in the second leader wild horse population and the second common wild horse population updated in the previous round of iteration is calculated in conjunction with the final mapping equation and the fitness function of the second wild horse population, and the position matrix of each leader wild horse and common wild horses in the second leader wild horse population and the second common wild horse population is updated again; S415, when , stop iteration and obtain the second final global optimal position; use the second final global optimal position as the adjusted first roughness data to be measured; when the adjusted first roughness data to be measured is less than the roughness threshold, the adjustment is completed; otherwise, return to S414 and continue to iterate until the adjusted first roughness data to be measured is less than the roughness threshold.

9. A system for implementing the intelligent surface roughness measurement method for motor rotor punchings as described in any one of claims 1 to 8.

Images