Method, medium and equipment for detecting service life of contact mechanism for molded case circuit breaker

Abstract

The invention discloses a service life detection method, medium and equipment for a contact mechanism of a molded case circuit breaker, and the method comprises the steps: obtaining a high-resolution spectrogram of the contact mechanism through solid nuclear magnetic resonance and magic angle rotation technologies, and carrying out the preprocessing of the high-resolution spectrogram to obtain reconstruction representation data; performing multi-step iterative prediction on the reconstructed data by using a multi-contact life decline prediction model so as to accurately obtain the life prediction condition of the contact mechanism; according to the method, reconstruction data are mapped to a high-dimensional feature space, an optimal separation hyperplane is constructed, an elastic net contact life SOH model is established based on the optimal separation hyperplane, and continuous scaling prediction is carried out. The method comprises the following steps: sorting and screening intrinsic characteristic variables of reconstructed data; and generating echelon life prediction of the contact mechanism by using a conditional generative adversarial network in combination with the reconstruction data and the SOH prediction condition. According to the method, service life detection of the contact mechanism is updated in real time, the detection precision is remarkably improved, prediction errors caused by parameter degradation are avoided, and a more accurate method is provided for service life prediction of the contact mechanism.

Description

Technical Field

[0001] The present invention relates to the technical field of molded case circuit breakers, and in particular to a contact mechanism life detection method, medium and equipment for molded case circuit breakers. Background Art

[0002] Molded case circuit breakers are important electrical equipment in low-voltage power distribution systems. The service life of molded case circuit breakers is related to the safe and reliable operation of low-voltage power distribution systems. Under the premise of no major structural failures, the service life of molded case circuit breakers mainly depends on the life of their contacts.

[0003] In the prior art, when predicting the life of the contacts of the molded case circuit breaker, the remaining life of the contacts is generally obtained by calculating the wear amount of the contact group.

[0004] However, in the existing implementation, since a major change is required in the structural design, the existing life prediction method of the circuit breaker contact has the problem of high prediction cost. Summary of the invention

[0005] 1. Technical issues to be resolved In order to solve the above problems, the present invention proposes a contact mechanism life detection method, medium and equipment for a molded case circuit breaker, aiming to solve the problem of high prediction cost in the life prediction method of the circuit breaker contact in the prior art.

[0006] (II) Technical solution A method for detecting the life of a contact mechanism for a molded case circuit breaker of the present invention comprises: Acquire characterization data of the contact mechanism in real time, and obtain reconstructed characterization data based on the characterization data; Based on a multiple contact life decay prediction model, a multi-step iterative prediction is performed on the reconstructed characterization data to obtain a life prediction of the contact mechanism; Converting the reconstructed representation data into an input vector, and mapping the input vector to a high-dimensional feature space, so as to construct an optimal separation hyperplane in the high-dimensional feature space; Based on the energy features in the reconstructed characterization data, an energy feature set is obtained, and an elastic net contact life SOH model is constructed based on an optimal separation hyperplane, and a continuous scaling prediction method is used to separate and predict the energy feature set to obtain the SOH prediction of the contact mechanism; Based on the reconstructed characterization data, the contact life parameter model of the contact mechanism is obtained, and the prediction relationship between the contact life parameter model and the SOH prediction situation is generated. The cascade life prediction of the contact mechanism is generated based on the correlation and the prediction relationship and the conditional generative adversarial network.

[0007] In the present invention, an empirical model of contact life attenuation is established, and a multi-step prediction model of contact life is established based on a Gaussian kernel function. The expression of the empirical model of contact life attenuation is: In the formula, express The contact life of the contact mechanism at all times, is the Coulomb efficiency, and Respectively represent the first determination parameter and the second determination parameter, is the initial moment, A constant representing time arrive The baseline effect of the time interval on the contact life, is the contact conductivity.

[0008] In the present invention, based on the contact life attenuation empirical model and the reconstructed characterization data, a one-step prediction of contact life degradation is performed on the contact mechanism. The expression of the contact life multi-step prediction model is: ; In the formula, express The predicted contact life at time represents the number of support vectors, represents the Lagrange multiplier, represents the Gaussian kernel function, defined as ,in is the Gaussian kernel parameter, yes The input vector of the reconstructed data at time instant, It's the amount of paranoia.

[0009] In the present invention, the expression of the one-step prediction model of the multi-step prediction model of contact life is: ; In the formula, represents the one-step forecast value, represents the multi-step prediction model of contact life, yes The delay vector at time ,in Represents the minimum dimension. The expression of the multi-step iterative prediction model of the contact life multi-step prediction model is: ; In the formula, Indicates from The time to start prediction The predicted value of the step, yes The input vector of the data is used to reconstruct the data at each moment, so that the resistance change rate of the contact mechanism can be accurately predicted.

[0010] In the present invention, the reconstructed characterization data is classified to obtain the temperature difference of the contact mechanism in the preset voltage range, the equal voltage difference in the closing state, and the fuzzy entropy based on the voltage; The temperature difference of the contact mechanism in a preset voltage range, the equal voltage difference in the closing state and the voltage-based fuzzy entropy are converted into corresponding input vectors to obtain several types of input vectors, and the several types of input vectors are mapped in a high-dimensional space.

[0011] In the present invention, an elastic net contact life SOH model is constructed in the optimal separation hyperplane, and the elastic net contact life SOH model is regularized to obtain a regularized model. Based on the regularized model and using a continuous scaling prediction method, the energy feature set is separated and predicted to obtain the optimal SOH prediction of the contact mechanism.

[0012] In the present invention, the SOH prediction is evaluated based on a linear regression model, and the expression of the parameter estimation value of the linear regression model is: ; In the formula, is the predicted value of SOH, is the energy feature set matrix, is the coefficient vector, and is a non-negative regularization parameter.

[0013] In the present invention, the linear correlation strength between the feature variables is assumed to be , based on the Pearson correlation coefficient, the following formula is obtained: in, and They are the sample values ​​of two feature variables. and is their mean.

[0014] Another computer-readable storage medium of the present invention stores a computer program thereon, which, when executed by a processor, implements the steps of the method for detecting the life of a contact mechanism for a molded case circuit breaker described in the above technical solution.

[0015] Another execution device of the present invention includes an execution device body and a controller, wherein the controller includes a processor and a computer program stored in a memory and executable on the processor, and when the processor executes the program, the steps of the contact mechanism life detection method for a molded case circuit breaker described in the above technical solution are implemented.

[0016] (III) Beneficial effects Compared with the prior art, the present invention has the following beneficial effects: (1) The present invention acquires the characterization data of the contact mechanism in real time and combines solid-state nuclear magnetic resonance and high-resolution magic angle spinning (HR-MAS) technology to deeply detect the internal environment of the contact mechanism and obtain high-resolution spectra, thus providing an accurate and detailed data basis for subsequent life prediction; The application of multiple contact life decay prediction models, especially the support vector machine model combined with the Gaussian kernel function, realizes the multi-step iterative prediction of the contact mechanism life. This method not only considers the current state, but also predicts the future state through iteration, which greatly improves the prediction accuracy and real-time performance.

[0017] (2) An elastic net contact life SOH model is constructed in the optimal separation hyperplane. By automatically selecting variables and continuously scaling, the problem of inconsistent degradation of contact mechanism parameters is effectively addressed, thus enhancing the robustness and generalization ability of the model. The regularized elastic net contact life SOH model further improves the stability and prediction accuracy of the model, ensuring that a high prediction accuracy can be maintained under complex and changeable working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 It is a logical structure diagram of the contact mechanism life detection method; Figure 2 Schematic diagram of the framework structure of the execution equipment.

[0020] 10. Processor, 11. Memory, 12. Communication interface, 13. Communication bus. DETAILED DESCRIPTION

[0021] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0022] Example 1

[0023] See also Figure 1 , which is a contact mechanism life detection method for a molded case circuit breaker in a first embodiment of the present invention, and the contact mechanism life detection method comprises steps S100 to S100: S100, acquiring the characterization data of the contact mechanism in real time, and preprocessing the characterization data to obtain reconstructed characterization data. Step S100 specifically includes: The solid-state nuclear magnetic resonance method is used in combination with a pulse sequence to detect the internal environment of the molded case circuit breaker, and based on the internal environment, a high-resolution magic angle spinning method (HR-MAS) is used to obtain a high-resolution spectrum of the contact mechanism. HR-MAS is developed based on the principle of magic angle spinning technology and is used to improve the resolution and sensitivity of the solid-state nuclear magnetic resonance spectrum. This is a prior art method and will not be elaborated on here.

[0024] It should be explained that the solid-state nuclear magnetic resonance method can be used to detect the internal environment in the contact mechanism, for example, the wear of the moving and static contacts in the contact mechanism and the ion chemical shift of the moving and static contacts in the contact mechanism when opening or closing the switch; in addition, the magic angle rotation method can eliminate the heteronuclear dipole coupling effect of the chemical shift anisotropy, so that high-resolution spectra of the moving and static contacts when opening and closing the switch can be obtained. For HR-MAS, the higher the speed used, the higher the resolution of the obtained spectrum. Solid-state nuclear magnetic resonance technology is based on the laws of quantum mechanics and uses the resonance absorption phenomenon produced by the atoms or nuclei of solids under a certain magnetic field. In the experiment, the sample is placed in a strong magnetic field, and the electrons and nuclei of the atoms in the sample are at different energy levels. If an alternating weak magnetic field of appropriate frequency is added in a plane perpendicular to the magnetic field to make some changes in the static magnetic field until resonance occurs, many atoms and nuclei will absorb energy from the resonant magnetic field and transition from low energy levels to high energy levels. At this time, the relationship between the sample's absorption of high-frequency electromagnetic energy and the magnetic field (or frequency) can be measured in the detection system, that is, the magnetic field resonance absorption curve. Based on the resonance line width, resonance absorption intensity and resonance curve shape of the magnetic resonance absorption curve, relevant information on the properties and structure of the sample can be obtained.

[0025] The background noise of the spectrum is eliminated based on the Hahn echo method, and the magic angle modulation phase sideband separation method is used to eliminate the rotation sideband signal in the spectrum; it should be explained that the principle of the Hahn echo is to add an additional focusing RF pulse on the basis of the magic angle rotation to reverse the spins with different precession rates. The Hahn echo method can be used to eliminate the background noise of part of the spectrum on the basis of the magic angle rotation, and the magic angle modulation phase sideband separation method can be used to eliminate the rotation sideband signal in the spectrum.

[0026] Furthermore, an intermittent opening and closing method is used to obtain the temperature difference of the contact mechanism in a preset voltage range, the equal voltage difference in the closed state, and the fuzzy entropy of the moving and static contacts based on the voltage; it can be understood that by performing opening and closing operations on the contact mechanism, the temperature difference of the contact mechanism in the preset voltage range and the temperature of the specified voltage range during the closing and opening process can be obtained, and then the equal voltage difference in the closed state of the contact mechanism and the fuzzy entropy of the moving and static contacts based on the voltage can be obtained.

[0027] The characterization data is sorted in chronological order, and the false neighbor method is used to determine the minimum dimension of spatial reconstruction; the characterization data sorted in time is reconstructed based on the minimum dimension to map the characterization data to the phase space to obtain the reconstructed characterization data. It should be explained that the temperature difference in the preset voltage range, the equal voltage difference in the closing state, and the fuzzy entropy based on the voltage of the moving and static contacts are sorted in chronological order, and then the false neighbor method is used to determine the minimum dimension of spatial reconstruction, and the first neighbor in the phase space is examined to determine the second neighbor in the phase space. Among them, the first neighbor is a real neighbor; the second neighbor is a false neighbor.

[0028] When there is no second neighboring point, the geometric structure can be considered to be fully opened. When two non-adjacent points in the high-dimensional attractor become adjacent points when projected onto the low-dimensional trajectory, such neighboring points are false. The proportion of false nearest neighboring points is calculated, and the dimension is gradually increased to obtain the minimum dimension. Then, the time-sorted characterization data is reconstructed according to the minimum dimension, that is, the temperature difference in the preset voltage range, the monomer internal resistance of the contact mechanism, the equal voltage difference in the closing state, and the voltage-based fuzzy entropy are reconstructed and mapped into the phase space to obtain the reconstructed characterization data of the contact mechanism.

[0029] S200, based on the multiple contact life decay prediction model, multi-step iterative prediction is performed on the reconstructed characterization data to obtain the life prediction of the contact mechanism. Step S200 specifically includes: Establish an empirical model of contact life attenuation and a multi-step prediction model of contact life based on Gaussian kernel function; In this embodiment, the expression of the contact life attenuation empirical model is: In the formula, express The contact life of the contact mechanism at all times, is the Coulomb efficiency, and Respectively represent the first determination parameter and the second determination parameter, is the initial moment, A constant representing time arrive The baseline effect of the time interval on the contact life, is the contact conductivity.

[0030] Based on the contact life attenuation empirical model and the reconstructed characterization data, a one-step prediction of the contact life degradation of the contact mechanism is performed; in this embodiment, the expression of the contact life multi-step prediction model is: In the formula, express The predicted contact life at time represents the number of support vectors, represents the Lagrange multiplier, represents the Gaussian kernel function, defined as ,in is the Gaussian kernel parameter, yes The input vector of the reconstructed data at time instant, It should be noted that the use of the data-driven model of the support vector machine to build a multi-step prediction model for contact life can effectively predict the regression model between the contact mechanism for a certain opening and closing cycle and the contact conductivity, and can provide observation values ​​for predicting the life of the contact mechanism.

[0031] Based on the multi-step prediction model of contact life and the reconstructed characterization data, a multi-step prediction is performed on the contact life of the contact mechanism to obtain the predicted resistance change rate of the contact mechanism; it can be understood that the use of multi-step iterative prediction can improve the prediction accuracy by using effective reconstructed data. In this embodiment, the expression of the one-step prediction model of the multi-step prediction model of contact life is: ; In the formula, represents the one-step forecast value, represents the multi-step prediction model of contact life, yes The delay vector at time ,in Represents the minimum dimension. The expression of the multi-step iterative prediction model of the contact life multi-step prediction model is: ; In the formula, Indicates from The time to start prediction The predicted value of the step, yes The input vector of the data is used to reconstruct the data at each moment, so that the resistance change rate of the contact mechanism can be accurately predicted.

[0032] S300, converting the reconstructed representation data into an input vector, and mapping the input vector to a high-dimensional feature space to construct an optimal separation hyperplane in the high-dimensional feature space. Step S300 includes: The reconstructed characterization data is classified to obtain the temperature difference of the contact mechanism in the preset voltage range, the equal voltage difference in the closing state, and the fuzzy entropy based on voltage. The temperature difference of the contact mechanism in the preset voltage range, the equal voltage difference in the closing state, and the fuzzy entropy based on voltage are converted into corresponding input vectors to obtain several types of input vectors, and the several types of input vectors are mapped in a high-dimensional space.

[0033] An optimal separation hyperplane is constructed based on several types of input vectors mapped in a high-dimensional space to distinguish the various characterization data in the contact mechanism; it can be understood that by classifying the reconstructed data of the contact mechanism, characterization data of different types of sodium ion contact life can be obtained, and the input vector is mapped to a high-dimensional feature space through nonlinear mapping, and an optimal separation hyperplane is constructed in the high-dimensional feature space, thereby converting the nonlinear problem in the low-dimensional space into a linear problem in the high-dimensional feature space, which can effectively distinguish the various characterization data in the contact mechanism.

[0034] S400, based on the energy features in the reconstructed characterization data, an energy feature set is obtained, and based on the optimal separation hyperplane, an elastic net contact life SOH model is constructed to separate and predict the energy feature set using a continuous scaling prediction method to obtain the SOH prediction of the contact mechanism. Step S400 specifically includes: An elastic net contact life SOH model is constructed in the optimal separation hyperplane, and the elastic net contact life SOH model is regularized to obtain a regularized model; Based on the regularization model and the continuous scaling prediction method, the energy feature set is separated and predicted to obtain the optimal SOH prediction of the contact mechanism; it can be understood that the various characterization data of the contact mechanism can be classified in the optimal separation hyperplane, and then the elastic net contact life SOH model is constructed in the optimal separation hyperplane.

[0035] In this embodiment, the elastic net contact life SOH model is a regularization process of linear regression, which can improve the prediction accuracy by automatically selecting variables and performing continuous scaling, so as to obtain the optimal SOH prediction of the contact mechanism, that is, the health of the contact life; through the energy characteristics of the contact life, such as the equal voltage difference in the closed state and the voltage-based fuzzy entropy, the goal of regularization optimization is usually to improve the SOH prediction results based on ordinary least squares by reducing parameter estimates to zero.

[0036] It should be explained that, in the specific implementation, given energy features, and then we can get the energy feature set , the equal voltage difference of the contact mechanism is named , The expression for the estimated value of is: ; In the formula, represents an estimate of the equal voltage difference of the contact mechanism, Represent the first energy feature, the second energy feature, and the Energy characteristics, They represent the estimated SOH fitting coefficient, the first fitting coefficient corresponding to the energy feature, the second fitting coefficient corresponding to the energy feature, and the third fitting coefficient corresponding to the energy feature. The fitted coefficients are, and the coefficient vector used to estimate SOH is , represents the number of fits, and the ordinary least squares method is used to minimize the residual The square sum of is used to estimate the coefficient vector; It is worth noting that the expression for the parameter estimates of the linear regression model used to estimate SOH is: ; In the formula, is the predicted value of SOH, is the energy feature set matrix, is the coefficient vector, and is a non-negative regularization parameter. The estimated value of the coefficient vector It can be obtained by the following formula: ,in, is the identity matrix, yes The symbol function of .

[0037] S500, sorting the intrinsic characteristic feature variables of the reconstructed characterization data, measuring the linear correlation strength between the sorted characteristic feature variables, and screening the sorted characteristic feature variables by selecting a preset threshold control to obtain the correlation between the monomer cycle life of the contact mechanism and the reconstructed characterization data.

[0038] Assume that the linear correlation strength between feature variables is , can be calculated by the Pearson correlation coefficient: in, and They are the sample values ​​of two feature variables. and is their mean.

[0039] Set the screening threshold to , only keep characteristic variables to ensure that the filtered data has a high degree of relevance.

[0040] It should be explained that, in this embodiment, the intrinsic characteristic feature variables of the reconstructed characterization data are the equal voltage difference in the closing state and the voltage-based fuzzy entropy. By sorting the equal voltage difference in the closing state and the voltage-based fuzzy entropy by time, the linear correlation strength between the equal voltage difference in the closing state and the voltage-based fuzzy entropy can be obtained, and then some data with low correlation can be filtered out by adopting a suitable filtering threshold, so that the filtered data and the linear correlation strength can effectively reflect the correlation between the single cycle life of the contact mechanism and these data.

[0041] S600, obtaining a contact life parameter model of the contact mechanism based on the reconstructed characterization data, generating a prediction relationship between the contact life parameter model and the SOH prediction situation, and generating a cascade life prediction of the contact mechanism based on the correlation and the prediction relationship and a conditional generative adversarial network.

[0042] It needs to be explained that the contact life parameter model of the contact mechanism is constructed according to the temperature difference of the contact mechanism in the preset voltage range, the equal voltage difference in the closing state and the fuzzy entropy based on the voltage. Since there is a strong nonlinear relationship between the various characterization data of the contact mechanism and the contact life SOH, the predictive relationship between the contact life parameter model of the contact mechanism and the SOH prediction situation is generated through the relationship between the various characterization data and the contact life SOH. That is, when the predictive relationship between the various characterization data of the contact mechanism and the contact life SOH prediction situation is generated, since the SOH prediction situation can provide a value indicator for the cascade utilization of the contact mechanism, the number of cycles that may continue when the contact mechanism is cascaded can be obtained; In this embodiment, the optimization objective function of the conditional generative adversarial network (CGAN) is: In the formula, is a generator, is the discriminator, It is real data. is a given label (i.e., various characterization data of the contact mechanism), is the input random noise signal, is the distribution of the real data, is the distribution of noise. By optimizing this objective function, a contact mechanism cascade life prediction similar to the real data distribution can be generated.

[0043] It can be understood that the real data in the original sample are the various characterization data of the contact mechanism. The use of conditional generative adversarial network can simulate the degradation characteristics of the contact mechanism in terms of correlation and prediction relationship, thereby enhancing the accuracy of the contact mechanism's cascade life prediction and obtaining a more accurate contact mechanism's cascade life prediction.

[0044] In summary, the contact mechanism life detection method in the above-mentioned embodiment of the present invention performs multi-step iterative prediction through multiple contact life decay prediction models, which can not only update the contact mechanism life detection in real time, but also effectively improve the accuracy of detection. By performing continuous scaling prediction on the elastic net contact life SOH model constructed in the optimal separation hyperplane, the reduction in prediction accuracy caused by inconsistent degradation of contact mechanism parameters can be avoided. The tiered life prediction of the contact mechanism generated by correlation and prediction relationship can more accurately predict the life of the contact mechanism.

[0045] Example 2

[0046] An embodiment of the present invention provides a computer-readable storage medium.

[0047] The computer-readable storage medium provided in the embodiment of the present invention stores a computer program, and when the computer program is executed by a processor, the steps of any one of the above-mentioned methods for detecting the life of a contact mechanism for a molded case circuit breaker can be implemented.

[0048] The computer-readable storage medium may include: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0049] For an introduction to the computer-readable storage medium provided in an embodiment of the present invention, please refer to the above method embodiment, and the present invention will not elaborate on it here.

[0050] The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0051] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0052] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0053] Example 3

[0054] An embodiment of the present invention provides an execution device.

[0055] Please refer to Figure 2 , Figure 2 This is a schematic diagram of the structure of an execution device provided by the present invention, and the execution device may include: Memory for storing computer programs; The processor can realize the steps of any one of the above-mentioned methods for detecting the life of the contact mechanism for a molded case circuit breaker when executing a computer program.

[0056] like Figure 2 , which is a schematic diagram of the composition structure of the execution device, the execution device may include: a processor 10, a memory 11, a communication interface 12 and a communication bus 13. The processor 10, the memory 11 and the communication interface 12 all communicate with each other through the communication bus 13.

[0057] In the embodiment of the present invention, the processor 10 may be a central processing unit (CPU), an application specific integrated circuit, a digital signal processor, a field programmable gate array or other programmable logic devices.

[0058] The processor 10 may call a program stored in the memory 11. Specifically, the processor 10 may execute the operations in the embodiment of the push button switch fault detection method.

[0059] The memory 11 is used to store one or more programs, which may include program codes, and the program codes include computer operation instructions. In the embodiment of the present invention, the memory 11 at least stores programs for implementing the following functions: Acquire characterization data of the contact mechanism in real time, and obtain reconstructed characterization data based on the characterization data; Based on a multiple contact life decay prediction model, a multi-step iterative prediction is performed on the reconstructed characterization data to obtain a life prediction of the contact mechanism; Converting the reconstructed representation data into an input vector, and mapping the input vector to a high-dimensional feature space, so as to construct an optimal separation hyperplane in the high-dimensional feature space; Based on the energy features in the reconstructed characterization data, an energy feature set is obtained, and an elastic net contact life SOH model is constructed based on an optimal separation hyperplane, and a continuous scaling prediction method is used to separate and predict the energy feature set to obtain the SOH prediction of the contact mechanism; Based on the reconstructed characterization data, the contact life parameter model of the contact mechanism is obtained, and the prediction relationship between the contact life parameter model and the SOH prediction situation is generated. The cascade life prediction of the contact mechanism is generated based on the correlation and the prediction relationship and the conditional generative adversarial network.

[0060] In a possible implementation, the memory 11 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and an application required for at least one function, etc.; the data storage area may store data created during use.

[0061] In addition, the memory 11 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device or other volatile solid-state storage device.

[0062] The communication interface 12 may be an interface of a communication module, and is used to connect to other devices or systems.

[0063] Of course, it should be noted that Figure 2 The structure shown does not constitute a limitation on the execution device in the embodiment of the present invention. In actual applications, the execution device may include Figure 2 More or fewer components than shown, or combinations of certain components.

[0064] The embodiments described above are merely descriptions of the preferred embodiments of the present invention, and are not intended to limit the concept and scope of the present invention. Without departing from the design concept of the present invention, various modifications and improvements made to the technical solution of the present invention by ordinary persons in the art should fall within the protection scope of the present invention, and the technical contents for which protection is requested by the present invention have been fully recorded in the claims. The embodiments described above are merely descriptions of the preferred embodiments of the present invention, and are not intended to limit the concept and scope of the present invention. Without departing from the design concept of the present invention, various modifications and improvements made to the technical solution of the present invention by ordinary persons in the art should fall within the protection scope of the present invention, and the technical contents for which protection is requested by the present invention have been fully recorded in the claims.

Claims

1. A method for detecting the life of a contact mechanism for a molded case circuit breaker, characterized in that: The contact mechanism life detection method for a molded case circuit breaker comprises: Acquire characterization data of the contact mechanism in real time, and obtain reconstructed characterization data based on the characterization data; Based on a multiple contact life decay prediction model, a multi-step iterative prediction is performed on the reconstructed characterization data to obtain a life prediction of the contact mechanism; Converting the reconstructed representation data into an input vector, and mapping the input vector to a high-dimensional feature space, so as to construct an optimal separation hyperplane in the high-dimensional feature space; Based on the energy features in the reconstructed characterization data, an energy feature set is obtained, and an elastic net contact life SOH model is constructed based on an optimal separation hyperplane, and a continuous scaling prediction method is used to separate and predict the energy feature set to obtain the SOH prediction of the contact mechanism; Based on the reconstructed characterization data, the contact life parameter model of the contact mechanism is obtained, and the prediction relationship between the contact life parameter model and the SOH prediction situation is generated. The cascade life prediction of the contact mechanism is generated based on the correlation and the prediction relationship and the conditional generative adversarial network.

2. The contact mechanism life detection method for molded case circuit breaker according to claim 1, characterized in that: An empirical model for contact life attenuation is established, and a multi-step prediction model for contact life is established based on a Gaussian kernel function. The expression of the empirical model for contact life attenuation is: In the formula, express The contact life of the contact mechanism at all times, is the Coulomb efficiency, and Respectively represent the first determination parameter and the second determination parameter, is the initial moment, A constant representing time arrive The baseline effect of the time interval on the contact life, is the contact conductivity.

3. The contact mechanism life detection method for molded case circuit breaker according to claim 2, characterized in that: Based on the contact life attenuation empirical model and the reconstructed characterization data, a one-step prediction of contact life degradation is performed on the contact mechanism. The expression of the contact life multi-step prediction model is: ; In the formula, express The contact life prediction value at time, represents the number of support vectors, represents the Lagrange multiplier, represents the Gaussian kernel function, defined as ,in is the Gaussian kernel parameter, yes The input vector of the reconstructed data at time instant, It's the amount of paranoia.

4. The contact mechanism life detection method for molded case circuit breaker according to claim 3, characterized in that: The expression of the one-step prediction model of the contact life multi-step prediction model is: ; In the formula, represents the one-step forecast value, represents the multi-step prediction model of contact life, yes The delay vector at time ,in Represents the minimum dimension. The expression of the multi-step iterative prediction model of the contact life multi-step prediction model is: ; In the formula, Indicates from The time to start prediction The predicted value of the step, yes The input vector of the data is used to reconstruct the data at each moment, so that the resistance change rate of the contact mechanism can be accurately predicted.

5. The contact mechanism life detection method for molded case circuit breaker according to claim 4, characterized in that: Classifying the reconstructed characterization data to obtain the temperature difference of the contact mechanism in a preset voltage range, the equal voltage difference in the closing state, and the fuzzy entropy based on voltage; The temperature difference of the contact mechanism in a preset voltage range, the equal voltage difference in the closing state and the voltage-based fuzzy entropy are converted into corresponding input vectors to obtain several types of input vectors, and the several types of input vectors are mapped in a high-dimensional space.

6. The contact mechanism life detection method for molded case circuit breaker according to claim 5, characterized in that: An elastic net contact life SOH model is constructed in the optimal separation hyperplane, and the elastic net contact life SOH model is regularized to obtain a regularized model. Based on the regularized model and using a continuous scaling prediction method, the energy feature set is separated and predicted to obtain the optimal SOH prediction of the contact mechanism.

7. The contact mechanism life detection method for molded case circuit breaker according to claim 6, characterized in that: The SOH prediction is evaluated based on a linear regression model, and the expression of the parameter estimation value of the linear regression model is: ; In the formula, is the predicted value of SOH, is the energy feature set matrix, is the coefficient vector, and is a non-negative regularization parameter.

8. The contact mechanism life detection method for molded case circuit breaker according to claim 7, characterized in that The linear correlation strength between the characteristic variables is , based on the Pearson correlation coefficient, the following formula is obtained: in, and They are the sample values ​​of two feature variables. and is their mean.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the contact mechanism life detection method for a molded case circuit breaker according to any one of claims 1 to 8 are implemented.

10. An execution device, comprising an execution device body and a controller, characterized in that: The controller includes a processor and a computer program stored in a memory and executable on the processor. When the processor executes the program, the steps of the method for detecting the life of a contact mechanism for a molded case circuit breaker according to any one of claims 1 to 8 are implemented.

Images