Motor insulation residual life prediction method and system

Abstract

The invention discloses a motor insulation residual life prediction method and system, and the method comprises the steps: building a motor insulation degradation model based on a Wiener process and an Arrhenius formula with residual breakdown voltage as a degradation amount, and building a motor insulation residual life prediction model based on the concept of first arrival; replacing later unknown temperature change with equivalent temperature, and setting model parameters as random variables which are in normal distribution and are mutually independent; joint prior distribution of the model parameters is obtained by using a maximum likelihood estimation method, and a posteriori distribution approximate solution of the model parameters is obtained by using Bayesian estimation and a Markov chain Monte Carlo method; and utilizing a total probability formula to obtain a final motor insulation residual life prediction model, and inputting the motor insulation degradation data of the motor winding under the time-varying temperature stress into the model to obtain the predicted motor insulation residual life. According to the method, the applicability and the flexibility are improved, and the life prediction result is more personalized and accurate.

Description

Technical Field

[0001] The present invention relates to the technical field of motor insulation remaining life prediction, and in particular to a motor insulation remaining life prediction method and system. Background Art

[0002] Motors are key equipment or components in the fields of power systems, new energy generation, motion control, etc. The reliability of motor insulation directly affects the stable operation of power or motion systems. During operation, due to the long-term effects of thermal, electrical, and mechanical stresses, the motor insulation material gradually ages and insulation breakdown may occur, which in turn affects the normal operation of the motor. Existing insulation life prediction methods are mainly based on product failure time statistical models and performance degradation models. However, in practical applications, the motor operating environment is complex and changeable, temperature changes have an important impact on insulation life, and there are differences between individual motors of the same type due to process, manufacturing and other factors. Therefore, it is necessary to study a method for predicting the remaining life of motor insulation that takes into account individual differences under time-varying temperature to improve the accuracy of the prediction. Summary of the invention

[0003] The purpose of the present invention is to provide a method and system for predicting the remaining life of motor insulation, establish a motor insulation remaining life prediction model based on the Wiener process and the Arrhenius formula, replace the unknown temperature changes with equivalent temperatures, and use the full probability formula to update the motor insulation remaining life prediction model to achieve high-precision remaining life prediction that is more in line with individual characteristics.

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

[0005] A method for predicting the remaining life of motor insulation comprises the following steps:

[0006] Taking the residual breakdown voltage as the degradation amount, according to the exponential relationship between the degradation rate and temperature reflected by the Arrhenius formula, a motor insulation remaining life prediction model based on the Wiener process and the Arrhenius formula was constructed under the concept of first arrival time; the model contains three parameters a, b, and c.

[0007] Collect the temperature of the motor and calculate the corresponding equivalent temperature.

[0008] Aiming at the individual differences of motor windings, that is, the parameters in the motor insulation remaining life prediction model of each motor winding are inconsistent, the parameters in the motor insulation remaining life prediction model are set as normally distributed and mutually independent random variables.

[0009] An accelerated aging test is performed on a set number of motor windings under different temperature stresses to obtain accelerated motor insulation degradation data of each motor winding, the data including residual breakdown voltage and time.

[0010] Based on the accelerated degradation data of motor insulation, the motor insulation remaining life prediction model is processed using the maximum likelihood estimation method to obtain the model parameter value corresponding to each motor winding. The obtained model parameter values ​​are processed again using the maximum likelihood estimation method to obtain the joint prior distribution of the model parameters.

[0011] Based on the equivalent temperature, the residual breakdown voltage data of the motor winding during operation under time-varying temperature stress is obtained. Combined with the data, the joint prior distribution of the model parameters is updated by using the Bayesian estimation and Markov chain Monte Carlo method, and the approximate solution of the posterior distribution of the model parameters is obtained.

[0012] Combined with the approximate solution of the posterior distribution of the model parameters and using the full probability formula, the final motor insulation remaining life prediction model is obtained. The residual breakdown voltage data of the motor winding during operation under time-varying temperature stress is input into the model to obtain the predicted motor insulation remaining life.

[0013] Furthermore, the construction of the motor insulation remaining life prediction model includes the following contents:

[0014] The specific expression of the degradation model is:

[0015]

[0016] Among them, U(t) represents the residual breakdown voltage value at the t-th operating time, U(0) represents the initial value of the residual breakdown voltage, B(t) represents the standard Brownian motion of randomness in the degradation process, and T represents the absolute temperature.

[0017] The expression of the motor insulation remaining life prediction model is:

[0018]

[0019] Where D represents the failure threshold of the remaining breakdown voltage, l represents the value of the remaining life, and f(l|t) represents the probability value that the remaining life is l at the tth operating time.

[0020] Furthermore, based on the principle of consistent degradation, the time-varying temperature in actual operation is discretized to obtain the equivalent temperature. The specific expression is:

[0021]

[0022] Where T0 represents the equivalent temperature, t urepresents the running time of the uth time period, u=1,2,…v, v represents the total number of time periods, T aclu Indicates t u The absolute temperature, t x Represents the sum of the running times of v time periods.

[0023] Furthermore, the parameters in the model are regarded as random variables that obey the normal distribution and are independent of each other. The specific expression is:

[0024]

[0025] Among them, p a 、p b 、p c Respectively represent the mean of model parameters a, b, c, q a ,q b ,q c Respectively represent the standard deviation of model parameters a, b, and c, Indicates that the mean is p a , standard deviation is The normal distribution of .

[0026] Further, obtaining the motor insulation accelerated degradation data of the motor winding includes the following contents:

[0027] Under the accelerated aging test, four insulation parameters of the motor winding, namely insulation resistance, insulation capacitance, dielectric loss factor and maximum partial discharge, are measured, and the accelerated degradation data of the motor insulation is obtained based on the four insulation parameters.

[0028] Furthermore, the prior distribution of the model parameters includes the following:

[0029] The likelihood function of the residual breakdown voltage change is:

[0030]

[0031] Use the likelihood function to find partial derivatives of the three model parameters respectively to obtain the model parameter value corresponding to each motor winding. The specific formula is:

[0032]

[0033] Where, L() represents the likelihood function of the residual breakdown voltage change, a ij represents the value of the model parameter a of the jth motor winding at the i-th temperature, and b ij represents the value of the model parameter b of the jth motor winding at the i-th temperature, c ijrepresents the value of the model parameter c of the jth motor winding at the i-th temperature, i = 1, 2 ... n1, n1 represents the total number of temperatures, j = 1, 2 ... n2, n2 represents the total number of motor windings, T i represents the absolute temperature of the ith temperature, Δt ijk represents the operating time increment between the residual breakdown voltage of the kth measurement and the k-1th measurement of the jth motor winding at the ith temperature, k = 1, 2…n3, n3 represents the total number of motor winding measurements, ΔU ijk It represents the reduction of the residual breakdown voltage value between the kth measurement and the k-1th measurement of the jth motor winding at the i-th temperature.

[0034] Combined with the parameter value a of each motor winding ij 、b ij 、c ij , the likelihood function of the model parameters is obtained using maximum likelihood estimation. The specific formula is:

[0035]

[0036] Use the likelihood function to find partial derivatives of the three model parameters respectively to obtain the mean and standard deviation of parameters a, b, and c. The specific formula is:

[0037]

[0038] Since the model parameters are independent of each other, the joint prior distribution of the model parameters is finally obtained as:

[0039] π0(a,b,c)=g(a)g(b)g(c)

[0040] Among them, π0(a,b,c) represents the prior distribution of model parameters, g(a), g(b), and g(c) represent the distribution of model parameters a, b, and c respectively.

[0041]

[0042] Furthermore, the approximate solution of the posterior distribution of the model parameters includes the following:

[0043] The prior distribution of model parameters is updated using Bayesian estimation to obtain the posterior distribution of model parameters. The specific expression is:

[0044]

[0045] Where π(a, b, c) represents the posterior distribution of the model parameters, ΔU represents the reduction in the remaining breakdown voltage, and Δt represents the increment of the running time.

[0046] The Markov Chain Monte Carlo method is used to sample from the expression of the posterior distribution of the model parameters to obtain the distribution of a, b, and c, which is used as the approximate solution of the posterior distribution of the model parameters.

[0047] Furthermore, the predicted remaining life of the motor insulation includes the following:

[0048] Combined with the approximate solution of the posterior distribution of the model parameters, the motor insulation remaining life prediction model is updated using the full probability formula to obtain the final motor insulation remaining life prediction model. The specific expression is:

[0049]

[0050] Wherein, U represents the residual breakdown voltage value.

[0051] The residual breakdown voltage data of the motor winding during operation under time-varying temperature stress is input into the model to obtain the predicted remaining life of the motor insulation. The specific expression is:

[0052]

[0053] in, represents the remaining life value of the t-th running time, E(l|U,t) represents the expected value of the remaining life, f(l|U,t) represents the probability density expression of the remaining life, and l represents the value of the remaining life.

[0054] Furthermore, the present invention also proposes a motor insulation remaining life prediction system, comprising:

[0055] The motor insulation remaining life prediction model construction module is used to construct a motor insulation remaining life prediction model based on the Wiener process and the Arrhenius formula under the concept of first arrival time, taking the remaining breakdown voltage as the degradation amount and according to the exponential relationship between the degradation rate and temperature reflected by the Arrhenius formula.

[0056] The equivalent temperature calculation module is used to collect the temperature of the motor and calculate the corresponding equivalent temperature.

[0057] The model parameter setting module is used to set the parameters in the motor insulation remaining life prediction model as normally distributed and mutually independent random variables in view of the individual differences of the motor windings, that is, the parameters in the motor insulation remaining life prediction model of each motor winding are inconsistent.

[0058] The motor insulation accelerated degradation data acquisition module is used to perform accelerated aging tests on a set number of motor windings under different temperature stresses to obtain the motor insulation accelerated degradation data of each motor winding, which includes the residual breakdown voltage and time.

[0059] The joint prior distribution acquisition module of the model parameters is used to process the motor insulation remaining life prediction model based on the motor insulation accelerated degradation data using the maximum likelihood estimation method to obtain the model parameter value corresponding to each motor winding, and the obtained model parameter value is processed again using the maximum likelihood estimation method to obtain the joint prior distribution of the model parameters.

[0060] The module for obtaining the approximate solution of the posterior distribution of model parameters is used to obtain the residual breakdown voltage data of the motor winding during operation under time-varying temperature stress based on the equivalent temperature. Combined with the data, the Bayesian estimation and Markov chain Monte Carlo method are used to update the joint prior distribution of the model parameters to obtain the approximate solution of the posterior distribution of the model parameters.

[0061] The motor insulation remaining life prediction module is used to combine the posterior distribution approximate solution of the model parameters and use the full probability formula to obtain the final motor insulation remaining life prediction model. The remaining breakdown voltage data of the motor winding during operation under time-varying temperature stress is input into the model to obtain the predicted motor insulation remaining life.

[0062] Furthermore, the present invention also proposes an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the motor insulation remaining life prediction method described above when executing the computer program.

[0063] Furthermore, the present invention also proposes a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is run by a processor, the motor insulation remaining life prediction method is executed.

[0064] Compared with the prior art, the present invention adopts the above technical solution and has the following technical effects:

[0065] 1. The method of the present invention can adapt to motors under different operating conditions, especially for time-varying temperature environments with complex and changeable working conditions, and improves applicability and flexibility by introducing equivalent temperature.

[0066] 2. The method of the present invention can better reflect the differences between different motors by treating the degradation model parameters as random variables and dynamically updating them in combination with individual degradation data, making the life prediction results more personalized and accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 It is an overall implementation flow chart of the present invention.

[0068] Figure 2 4 is a graph showing the residual breakdown voltage data at high temperature of an embodiment of the present invention.

[0069] Figure 3 4 is a graph showing residual breakdown voltage data under time-varying temperature according to an embodiment of the present invention.

[0070] Figure 4 3 is a probability density curve diagram of the remaining life of the motor insulation at different operating times according to an embodiment of the present invention.

[0071] Figure 5 It is a schematic diagram of the prediction results of the remaining insulation life of a motor under different operating times according to an embodiment of the present invention. DETAILED DESCRIPTION

[0072] The present invention will be further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the protection scope of the present invention.

[0073] To achieve the above object, the present invention proposes a method for predicting the remaining life of motor insulation, such as Figure 1 The specific contents are as follows:

[0074] Taking the residual breakdown voltage as the degradation amount, according to the exponential relationship between the degradation rate and temperature reflected by the Arrhenius formula, a motor insulation residual life prediction model based on the Wiener process and the Arrhenius formula is constructed under the concept of first arrival time. The model contains three parameters a, b, and c; the specific contents are:

[0075] The specific expression of the degradation model is:

[0076] U(t)=U(0)-μt-σB(t) (1)

[0077] Among them, U(t) represents the residual breakdown voltage value at the t-th operating time; U(0) represents the initial value of the residual breakdown voltage; B(t) represents the standard Brownian motion of randomness in the degradation process; μ is the drift coefficient, which represents the degradation rate; σ is the diffusion coefficient, which represents the influence of random factors in the degradation process.

[0078] Under the action of thermal stress, the drift coefficient and diffusion coefficient satisfy the Arrhenius formula, and the specific expression is:

[0079]

[0080] Here, T represents absolute temperature.

[0081] Substituting formula (2) into formula (1), the motor insulation degradation model is finally established as:

[0082]

[0083] The expression of the motor insulation remaining life prediction model is:

[0084]

[0085] Where D represents the failure threshold of the remaining breakdown voltage, l represents the value of the remaining life, and f(l|t) represents the probability value that the remaining life is l at the tth operating time.

[0086] From the characteristics of the Wiener process, we can know that:

[0087]

[0088] Where ΔU represents the reduction in the residual breakdown voltage and Δt represents the increment in the operating time.

[0089] Since the temperature change of the motor is unknown, the time-varying temperature is converted into a constant equivalent temperature to replace the subsequent unknown temperature. Therefore, the temperature of the motor is collected and the corresponding equivalent temperature is calculated for subsequent remaining life prediction; the specific contents are as follows:

[0090] Based on the principle of consistent degradation, the time-varying temperature in actual operation is discretized to obtain the equivalent temperature. The specific expression is:

[0091]

[0092] Where T0 represents the equivalent temperature, t u represents the running time of the uth time period, u=1,2,…v, v represents the total number of time periods, T aclu Indicates t u The absolute temperature, t x Represents the sum of the running times of v time periods.

[0093] In view of the individual differences of motor windings, that is, the parameters in the motor insulation remaining life prediction model of each motor winding are inconsistent, the parameters in the motor insulation remaining life prediction model are set as normally distributed and mutually independent random variables. The specific expression is:

[0094]

[0095] Among them, p a 、p b 、p c Respectively represent the mean of model parameters a, b, c, q a ,q b ,q c Respectively represent the standard deviation of model parameters a, b, and c, Indicates that the mean is p a , standard deviation is The normal distribution of .

[0096] Perform accelerated aging tests on a set number of motor windings under different temperature stresses to obtain accelerated motor insulation degradation data for each motor winding, including residual breakdown voltage and time; the specific contents are:

[0097] Under the accelerated aging test, four insulation parameters of the motor winding, namely insulation resistance, insulation capacitance, dielectric loss factor and maximum partial discharge, are measured, and the accelerated degradation data of the motor insulation is obtained based on the four insulation parameters.

[0098] Based on the accelerated degradation data of motor insulation, the motor insulation remaining life prediction model is processed using the maximum likelihood estimation method to obtain the model parameter value corresponding to each motor winding. The obtained model parameter value is processed again using the maximum likelihood estimation method to obtain the joint prior distribution of the model parameters; the specific content is:

[0099] According to formula (5), combined with Figure 2 The maximum likelihood estimate is solved for the data in , and the likelihood function of the residual breakdown voltage change is:

[0100]

[0101] Use the likelihood function to find partial derivatives of the three model parameters respectively to obtain the model parameter value corresponding to each motor winding. The specific formula is:

[0102]

[0103] Where, L() represents the likelihood function of the residual breakdown voltage change, a ij represents the value of the model parameter a of the jth motor winding at the i-th temperature, and b ij represents the value of the model parameter b of the jth motor winding at the i-th temperature, c ij represents the value of the model parameter c of the jth motor winding at the i-th temperature, i = 1, 2 ... n1, n1 represents the total number of temperatures, j = 1, 2 ... n2, n2 represents the total number of motor windings, T i represents the absolute temperature of the ith temperature, Δt ijk represents the operating time increment between the residual breakdown voltage of the kth measurement and the k-1th measurement of the jth motor winding at the ith temperature, k = 1, 2…n3, n3 represents the total number of motor winding measurements, ΔU ijk It represents the reduction in the residual breakdown voltage value between the kth measurement and the k-1th measurement of the jth motor winding at the ith temperature.

[0104] Combined with the parameter value a of each motor winding ij , b ij 、c ij, the likelihood function of the model parameters is obtained using maximum likelihood estimation. The specific formula is:

[0105]

[0106] Use the likelihood function to find partial derivatives of the three model parameters respectively to obtain the mean and standard deviation of parameters a, b, and c. The specific formula is:

[0107]

[0108] Since the model parameters are independent of each other, the joint prior distribution of the model parameters is finally obtained as:

[0109] π0(a,b,c)=g(a)g(b)g(c)(12)

[0110] Among them, π0(a,b,c) represents the prior distribution of model parameters, g(a), g(b), and g(c) represent the distribution of model parameters a, b, and c respectively.

[0111] Based on the equivalent temperature, the residual breakdown voltage data of the motor winding under time-varying temperature stress in actual operation is obtained. Combined with this data, the Bayesian estimation and Markov chain Monte Carlo method are used to update the joint prior distribution of the model parameters, and the approximate solution of the posterior distribution of the model parameters is obtained; the specific contents are as follows:

[0112] The prior distribution of model parameters is updated using Bayesian estimation to obtain the posterior distribution of model parameters. The specific expression is:

[0113]

[0114] Among them, π(a,b,c) represents the posterior distribution of model parameters.

[0115] Since the model parameters are relatively complex, the Markov Chain Monte Carlo method is used to sample from the expression of the posterior distribution of the model parameters to obtain the distribution of a, b, and c, which is used as the approximate solution of the posterior distribution of the model parameters.

[0116] Combined with the approximate solution of the posterior distribution of the model parameters, the final motor insulation remaining life prediction model is obtained using the full probability formula. The residual breakdown voltage data of the motor winding during actual operation under time-varying temperature stress is input into the model to obtain the predicted motor insulation remaining life; the specific contents are as follows:

[0117] Combined with the approximate solution of the posterior distribution of the model parameters, the motor insulation remaining life prediction model is updated using the full probability formula to obtain the final motor insulation remaining life prediction model. The specific expression is:

[0118]

[0119] Wherein, U represents the residual breakdown voltage value.

[0120] The residual breakdown voltage data of the motor winding during actual operation under time-varying temperature stress is input into the model to obtain the predicted remaining life of the motor insulation. The specific expression is:

[0121]

[0122] in, represents the remaining life value of the t-th running time, E(l|U,t) represents the expected value of the remaining life, f(l|U,t) represents the probability density expression of the remaining life, and l represents the value of the remaining life.

[0123] This embodiment is simulated and tested in Matlab. Figure 2 (a), (b), (c) and (d) are the residual breakdown voltage data at 180℃, 200℃, 220℃ and 240℃ respectively. There are degradation data of 5 motor windings at each temperature. The residual breakdown voltage data is obtained every 50 hours until 500 hours. The initial value of the residual breakdown voltage is 40kV. As time increases, the residual breakdown voltage gradually decreases.

[0124] Figure 3 The data in the figure are the residual breakdown voltage data under time-varying temperature. During operation, the temperature changes periodically at 80°C, 100°C, and 120°C. The operation time at each temperature is distributed in a ratio of 5:3:2 with a cycle of 250 hours until failure. Figure 3 The model parameters are updated every 800 hours based on the data, and the remaining life prediction is performed.

[0125] Figure 4 and Figure 5 This is the prediction result of the remaining life of the motor insulation. Figure 4 As shown in the figure, with the increase of running time, the probability density curve of the remaining life gradually narrows, which shows that the uncertainty of the prediction results is getting lower and lower. On the one hand, the increase in the amount of degradation data obtained makes the prediction results more accurate, and on the other hand, with the increase of running time, the equivalent temperature is more consistent with the change of actual temperature. Figure 5 As shown in the figure, with the increase of running time, after updating the model parameters in combination with the motor winding degradation data, the predicted value of the remaining life will be closer to the actual value, and a high-precision remaining life prediction that is more in line with the characteristics of the motor winding can be achieved.

[0126] The embodiment of the present invention also proposes a motor insulation remaining life prediction system, including a motor insulation remaining life prediction model building module, an equivalent temperature calculation module, a model parameter setting module, a motor insulation accelerated degradation data acquisition module, a joint prior distribution acquisition module of model parameters, a posterior distribution approximate solution acquisition module of model parameters, a motor insulation remaining life prediction module and a computer program that can be run on a processor. It should be noted that each module in the above system corresponds to the specific steps of the method provided in the embodiment of the present invention, and has the functional modules and beneficial effects corresponding to the execution method. For technical details not fully described in this embodiment, please refer to the method provided in the embodiment of the present invention.

[0127] The embodiment of the present invention further provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. It should be noted that when the processor executes the computer program, the specific steps of the method provided in the embodiment of the present invention are corresponding to the specific steps of the method provided in the embodiment of the present invention, and the processor has the functional modules and beneficial effects corresponding to the execution method. For technical details not described in detail in this embodiment, please refer to the method provided in the embodiment of the present invention.

[0128] The embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program. It should be noted that when the computer program is executed by the processor, it corresponds to the specific steps of the method provided in the embodiment of the present invention, and has the functional modules and beneficial effects corresponding to the execution method. For technical details not described in detail in this embodiment, please refer to the method provided in the embodiment of the present invention.

[0129] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for predicting the remaining life of motor insulation, characterized in that: include: Taking the residual breakdown voltage as the degradation amount, according to the exponential relationship between the degradation rate and temperature reflected by the Arrhenius formula, a motor insulation residual life prediction model based on the Wiener process and the Arrhenius formula was constructed under the concept of first arrival time. Collect the temperature of the motor and calculate the corresponding equivalent temperature; The parameters in the motor insulation remaining life prediction model are set as normally distributed and mutually independent random variables; Performing accelerated aging tests on a set number of motor windings under different temperature stresses to obtain accelerated motor insulation degradation data of each motor winding, the data including residual breakdown voltage and time; Based on the accelerated degradation data of motor insulation, the motor insulation remaining life prediction model is processed using the maximum likelihood estimation method to obtain the model parameter value corresponding to each motor winding. The obtained model parameter value is processed again using the maximum likelihood estimation method to obtain the joint prior distribution of the model parameters. Based on the equivalent temperature, the residual breakdown voltage data of the motor winding under time-varying temperature stress during operation is obtained. Combining this data, the Bayesian estimation and Markov chain Monte Carlo method are used to update the joint prior distribution of the model parameters, and the approximate solution of the posterior distribution of the model parameters is obtained. Combined with the approximate solution of the posterior distribution of the model parameters and using the full probability formula, the final motor insulation remaining life prediction model is obtained. The residual breakdown voltage data of the motor winding during operation under time-varying temperature stress is input into the model to obtain the predicted motor insulation remaining life.

2. The method for predicting the remaining life of motor insulation according to claim 1, characterized in that: The construction of the motor insulation remaining life prediction model includes the following: The specific expression of the degradation model is: Where U(t) represents the residual breakdown voltage value at the t-th operating time, U(0) represents the initial value of the residual breakdown voltage, B(t) represents the standard Brownian motion of randomness in the degradation process, T represents the absolute temperature, and a, b, and c all represent model parameters; The expression of the motor insulation remaining life prediction model is: Where D represents the failure threshold of the remaining breakdown voltage, l represents the value of the remaining life, and f(l|t) represents the probability value that the remaining life is l at the tth operating time.

3. The method for predicting the remaining life of motor insulation according to claim 1, characterized in that: Based on the principle of consistent degradation, the time-varying temperature is discretized to obtain the equivalent temperature. The specific expression is: Where T0 represents the equivalent temperature, t u represents the running time of the uth time period, u=1,2,…v, v represents the total number of time periods, T aclu Indicates t u The absolute temperature, t x represents the sum of the running time of v time periods, and b represents the model parameters.

4. The method for predicting the remaining life of motor insulation according to claim 1, characterized in that: The specific expressions of the model parameters are: Among them, p a 、p b 、p c Respectively represent the mean of model parameters a, b, c, q a ,q b ,q c They represent the standard deviations of the model parameters a, b, and c respectively, and N( ) represents the normal distribution.

5. The method for predicting the remaining life of motor insulation according to claim 1, characterized in that: Obtaining accelerated motor insulation degradation data for motor windings involves the following: Under the accelerated aging test, four insulation parameters of the motor winding, namely insulation resistance, insulation capacitance, dielectric loss factor and maximum partial discharge, are measured, and the accelerated degradation data of the motor insulation is obtained based on the four insulation parameters.

6. The method for predicting the remaining life of motor insulation according to claim 1, characterized in that: The prior distribution of the model parameters includes the following: The likelihood function of the residual breakdown voltage change is: Use the likelihood function to find partial derivatives of the three model parameters respectively to obtain the model parameter value corresponding to each motor winding. The specific formula is: Where, L() represents the likelihood function of the residual breakdown voltage change, a ij represents the value of the model parameter a of the jth motor winding at the i-th temperature, and b ij represents the value of the model parameter b of the jth motor winding at the i-th temperature, c ij represents the value of the model parameter c of the jth motor winding at the i-th temperature, i = 1, 2 ... n1, n1 represents the total number of temperatures, j = 1, 2 ... n2, n2 represents the total number of motor windings, T i represents the absolute temperature of the ith temperature, Δt ijk represents the operating time increment between the residual breakdown voltage of the kth measurement and the k-1th measurement of the jth motor winding at the ith temperature, k = 1, 2…n3, n3 represents the total number of motor winding measurements, ΔU ijk represents the reduction in the residual breakdown voltage value between the kth measurement and the k-1th measurement of the jth motor winding at the i-th temperature, and a, b, and c all represent model parameters; Combined with the parameter value a of each motor winding ij , b ij 、c ij , the likelihood function of the model parameters is obtained using maximum likelihood estimation. The specific formula is: Use the likelihood function to find partial derivatives of the three model parameters respectively to obtain the mean and standard deviation of parameters a, b, and c. The specific formula is: Among them, p a 、p b 、p c Respectively represent the mean of model parameters a, b, c, q a ,q b ,q c represent the standard deviations of model parameters a, b, and c, respectively; Since the model parameters are independent of each other, the joint prior distribution of the model parameters is finally obtained as: π0(a,b,c)=g(a)g(b)g(c) Among them, π0(a,b,c) represents the prior distribution of model parameters, g(a), g(b), and g(c) represent the distribution of model parameters a, b, and c respectively.

7. The method for predicting the remaining life of motor insulation according to claim 1, characterized in that: The approximate solution to the posterior distribution of the model parameters includes the following: The prior distribution of model parameters is updated using Bayesian estimation to obtain the posterior distribution of model parameters. The specific expression is: Among them, π(a, b, c) represents the posterior distribution of model parameters, ΔU represents the reduction of residual breakdown voltage, Δt represents the increment of running time, L() represents the likelihood function of the change of residual breakdown voltage, π0(a, b, c) represents the prior distribution of model parameters, and a, b, c all represent model parameters; The Markov Chain Monte Carlo method is used to sample from the expression of the posterior distribution of the model parameters to obtain the distribution of a, b, and c, which is used as the approximate solution of the posterior distribution of the model parameters.

8. The method for predicting the remaining life of motor insulation according to claim 1, characterized in that: The predicted remaining life of the motor insulation includes the following: Combined with the approximate solution of the posterior distribution of the model parameters, the motor insulation remaining life prediction model is updated using the full probability formula to obtain the final motor insulation remaining life prediction model. The specific expression is: Where U represents the residual breakdown voltage value, π(a, b, c) represents the posterior distribution of the model parameters, t represents the running time, and a, b, and c all represent model parameters; The residual breakdown voltage data of the motor winding during operation under time-varying temperature stress is input into the model to obtain the predicted remaining life of the motor insulation. The specific expression is: in, represents the remaining life value of the t-th running time, E(l|U,t) represents the expected value of the remaining life, f(l|U,t) represents the probability density expression of the remaining life, and l represents the value of the remaining life.

9. A system for predicting the remaining life of motor insulation according to any one of claims 1 to 8, characterized in that: include: The module for building a motor insulation remaining life prediction model is used to build a motor insulation remaining life prediction model based on the Wiener process and the Arrhenius formula under the concept of first arrival time, taking the remaining breakdown voltage as the degradation amount and according to the exponential relationship between the degradation rate and temperature reflected by the Arrhenius formula; The equivalent temperature calculation module is used to collect the temperature of the motor and calculate the corresponding equivalent temperature; A model parameter setting module, used to set the parameters in the motor insulation remaining life prediction model as normally distributed and mutually independent random variables; A motor insulation accelerated degradation data acquisition module is used to perform accelerated aging tests on a set number of motor windings under different temperature stresses to obtain motor insulation accelerated degradation data of each motor winding, the data including residual breakdown voltage and time; A joint prior distribution acquisition module for model parameters is used to process the motor insulation remaining life prediction model based on the motor insulation accelerated degradation data using the maximum likelihood estimation method to obtain the model parameter value corresponding to each motor winding, and the obtained model parameter value is processed again using the maximum likelihood estimation method to obtain the joint prior distribution of the model parameters; The module for obtaining the approximate solution of the posterior distribution of the model parameters is used to obtain the residual breakdown voltage data of the motor winding during operation under time-varying temperature stress based on the equivalent temperature. Based on the data, the Bayesian estimation and Markov chain Monte Carlo method are used to update the joint prior distribution of the model parameters to obtain the approximate solution of the posterior distribution of the model parameters. The motor insulation remaining life prediction module is used to combine the posterior distribution approximate solution of the model parameters and use the full probability formula to obtain the final motor insulation remaining life prediction model. The remaining breakdown voltage data of the motor winding during operation under time-varying temperature stress is input into the model to obtain the predicted motor insulation remaining life.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method for predicting the remaining life of motor insulation as claimed in any one of claims 1 to 8 are implemented.

11. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for predicting the remaining life of motor insulation according to any one of claims 1 to 8 is executed.

Images