SVM-based substation equipment fault diagnosis method and system, and medium

Abstract

The invention relates to a substation equipment fault diagnosis method and system based on an SVM and a medium, and the method comprises the steps: S1, obtaining the monitoring data of substation equipment, carrying out the nonlinear conversion through combining with the SVM, and determining a target function; s2, based on ICA, SVM parameter optimization is carried out on the objective function; and S3, after SVM parameter optimization, performing fault diagnosis on the substation equipment based on a nonlinear SVM. SVM parameter optimization is carried out on the objective function through ICA, so that fault diagnosis is carried out on the substation equipment through a nonlinear SVM, and the accuracy and reliability of fault diagnosis of the substation equipment are further improved.

Description

Technical Field

[0001] The embodiments of the present invention relate to the technical field of substation abnormality detection, and in particular to a substation equipment fault diagnosis method, system and medium based on SVM. Background Art

[0002] Anomaly detection of smart substation equipment is a key technology that is of great significance in identifying and predicting equipment or system failures.

[0003] Traditional anomaly detection methods are usually passive, triggering alerts only when indicators exceed preset thresholds or deviate significantly from historical data. However, this approach may not accurately predict extreme data, thus reducing the accuracy of anomaly detection. Summary of the invention

[0004] This application is made to solve the above problems, and its purpose is to provide a substation equipment fault diagnosis method, system and medium based on SVM to improve the accuracy of substation equipment fault diagnosis. In order to achieve the above purpose, this application adopts the following scheme:

[0005] The present application provides a substation equipment fault diagnosis method based on SVM, and the substation equipment fault diagnosis method based on SVM comprises the following steps:

[0006] S1. Obtain monitoring data of substation equipment, perform nonlinear transformation in combination with SVM, and determine the objective function;

[0007] S2. Optimizing SVM parameters of the objective function based on ICA;

[0008] S3. After the SVM parameters are optimized, fault diagnosis of the substation equipment is performed based on the nonlinear SVM;

[0009] Among them, the objective function satisfies the following relationship:

[0010]

[0011] Among them, f(x) is the objective function, α i is the horizontal intercept of the ith statistic of the substation equipment, b is the vertical intercept of the statistic of the substation equipment, y i is the index of the i-th statistic, K(x,x i ) is the fault threshold function of the substation equipment, x i is the parameter of the i-th statistic.

[0012] Further, the monitoring data of the substation equipment is obtained, and the optimization target is determined to determine the objective function, wherein the optimization target satisfies the following relationship:

[0013]

[0014] Wherein, ω is the failure cycle statistic of the substation equipment, C is the factor of the substation equipment, ξ i is the fault degree index of the i-th substation equipment, is the fault impact of the i-th substation equipment.

[0015] Furthermore, after determining the optimization target, the optimization target is dualized, and the dualized form satisfies the following relationship:

[0016]

[0017] Among them, Ψ(α) is the dual maximum function, y j is the index of the jth statistic, and l is the number of indicators of the statistic.

[0018] Further, S2, based on ICA, the step of performing SVM parameter optimization on the objective function specifically includes:

[0019] Randomly select some points on the function and initialize the empire;

[0020] the removal of colonies to relevant imperialist countries;

[0021] If there is a colony in the empire that costs less than the imperialist, swap the positions of the colony and the imperialist;

[0022] Calculate the total cost of all empires;

[0023] Take the weakest colony from the weakest empire and give it to the empire most likely to have him, and destroy the weak empire.

[0024] Further, after S3, SVM parameters are optimized, the step of performing fault diagnosis on the substation equipment based on nonlinear SVM specifically includes:

[0025] Use PCA to perform linear transformation on the initial sample set, use the cumulative contribution rate after transformation to select the principal component, and perform dimensionality reduction processing on the sample set after linear transformation based on the selected principal component to obtain a new sample set;

[0026] Centrally training the new samples, wherein an objective function is formed by the new sample set and a multi-classification support vector machine with unknown parameters;

[0027] Use ICA to find the optimal solution of the objective function to determine the optimal parameters of SVM for smart substation fault classification;

[0028] Input the test sample into the support vector machine model and calculate its predicted value.

[0029] Furthermore, when using PCA to perform linear transformation on the initial sample set, the RBF kernel function is selected as the kernel function of the training model.

[0030] The present application also provides a substation equipment fault diagnosis system based on SVM, and the substation equipment fault diagnosis system based on SVM includes:

[0031] The objective function determination module is configured to obtain monitoring data of the substation equipment, perform nonlinear transformation in combination with the SVM, and determine the objective function;

[0032] A parameter optimization module is configured to perform SVM parameter optimization on the objective function based on ICA;

[0033] The fault diagnosis module is configured to perform fault diagnosis on the substation equipment based on nonlinear SVM after SVM parameter optimization.

[0034] Furthermore, the substation equipment fault diagnosis system based on SVM includes:

[0035] The linear transformation module is configured to perform linear transformation on the initial sample set through PCA, select the principal component using the cumulative contribution rate after the transformation, and perform dimensionality reduction processing on the sample set after the linear transformation based on the selected principal component.

[0036] The present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the substation equipment fault diagnosis method based on SVM as described in any one of the above items is implemented.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] By combining SVM for nonlinear transformation, this application can more accurately identify abnormal patterns of substation equipment and effectively improve the accuracy of anomaly detection even in the case of complex data distribution. Compared with traditional threshold triggering methods, this application can better handle extreme data and potential nonlinear relationships, thereby improving the ability to predict equipment or system failures. The SVM parameters are optimized through independent component analysis (ICA), and the model parameters are automatically adjusted to adapt to different substation equipment operating conditions, thereby improving the adaptability of the diagnostic method.

[0039] This application has made significant breakthroughs in the field of substation equipment fault analysis, improved the accuracy of fault diagnosis and prediction, is suitable for processing complex data relationships, and improves the operating efficiency of substation equipment fault analysis, with significant practical value and economic benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. 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 creative work.

[0041] Figure 1 is a flow chart of substation equipment fault diagnosis based on SVM involved in an embodiment of the present application;

[0042] Figure 2 It is a schematic diagram of substation equipment fault diagnosis based on SVM involved in an embodiment of the present application. DETAILED DESCRIPTION

[0043] The specific implementation methods of the present application are described in detail below with reference to the accompanying drawings.

[0044] It should be noted that, in the absence of conflict, the embodiments and technical features in the embodiments of the present application can be combined with each other, and the detailed description in the specific implementation method should be understood as an explanation of the purpose of the present application and should not be regarded as an improper limitation on the present application.

[0045] It should be understood that the orientation or position relationship is based on the orientation or position relationship shown in the drawings. These orientation terms are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present application.

[0046] Fault analysis technology for smart substation equipment is a key technology that is of great significance in identifying and predicting equipment or system failures. The effective implementation of this technology requires multiple steps, from abnormality detection to fault diagnosis and then to fault prediction. The following is a brief introduction to these steps:

[0047] Anomaly detection: First, the various indicators of the equipment or system need to be monitored and detected to determine whether they are abnormal. These indicators may cover parameters such as temperature, pressure, current, voltage, etc. Anomalies can be indicators exceeding the preset threshold or showing significant deviations compared with historical data.

[0048] Feature extraction: Once an anomaly is found, the next step is to extract features from these abnormal indicators. These features may include statistics such as the mean, variance, and rate of change of the indicator, or more complex features such as spectral features, time-frequency features, etc.

[0049] Model training: The extracted features will be used as input to train the machine learning model. One of the commonly used methods here is support vector machine, which can detect anomalies in new data by learning the characteristic patterns under normal operating conditions.

[0050] Fault diagnosis: The trained model can be used to classify new data to determine whether the device or system is in a normal state or has a fault. If the model identifies certain features that match known failure modes, further fault diagnosis and treatment can be performed.

[0051] Through the above steps, fault analysis technology can help detect potential faults of smart substation equipment in advance and take corresponding maintenance or repair measures to ensure its normal operation and reliability.

[0052] However, existing anomaly detection methods are usually passive and can only trigger alerts when indicators exceed preset thresholds or show significant deviations compared with historical data.

[0053] In view of this, the embodiment of the present application proposes a substation equipment fault diagnosis method based on SVM, which is based on ICA and SVM. Figure 1 is a flow chart of substation equipment fault diagnosis based on SVM involved in an embodiment of the present application, Figure 2 FIG. 1 is a schematic diagram of a substation equipment fault diagnosis based on SVM according to an embodiment of the present application. Figure 1 and Figure 2 As shown, the substation equipment abnormality detection method includes the following steps:

[0054] S1. Obtain monitoring data of substation equipment, perform nonlinear transformation in combination with SVM, and determine the objective function;

[0055] S2. Optimizing SVM parameters of the objective function based on ICA;

[0056] S3. After the SVM parameters are optimized, fault diagnosis of the substation equipment is performed based on the nonlinear SVM;

[0057] Among them, the objective function satisfies the following relationship:

[0058]

[0059] Among them, f(x) is the objective function, α i is the horizontal intercept of the ith statistic of the substation equipment, b is the vertical intercept of the statistic of the substation equipment, y i is the index of the i-th statistic, K(x,x i ) is the fault threshold function of the substation equipment, x i is the parameter of the i-th statistic.

[0060] Specifically, the monitoring data of the equipment, including key indicators such as current, voltage, temperature, vibration, etc., are obtained in real time from the SCADA system of the substation. The support vector machine (SVM) is used to perform nonlinear transformation on these monitoring data to extract features and construct the objective function. For example, a kernel function such as the radial basis function (RBF) can be used to process nonlinear data. The form of the objective function is shown in the above formula (1.1). The objective function is decomposed using the independent component analysis (ICA) technique to find the independent components between the statistics, so as to better understand the intrinsic structure of the data. According to the results of ICA, the parameters of the SVM are optimized, including the parameters of the kernel function, the penalty parameter C, etc., to maximize the accuracy of fault diagnosis. For example, through ICA analysis, it may be found that some statistics contribute more to fault diagnosis, so the weights wi of these statistics can be adjusted.

[0061] After the SVM parameters are optimized, the substation equipment is diagnosed for faults based on the nonlinear SVM. After the SVM parameter optimization is completed, the nonlinear SVM model is used to diagnose the substation equipment for faults. For new monitoring data, it is input into the trained nonlinear SVM model. The model will output a diagnosis result indicating whether the equipment is in a normal state or has a fault. For example, if the diagnosis result shows that F(x) exceeds the fault threshold function, the system will trigger an alarm, indicating that the equipment may have a fault.

[0062] It should be noted that by combining SVM for nonlinear transformation, the abnormal patterns of substation equipment can be identified more accurately, and the accuracy of anomaly detection can be effectively improved even in the case of complex data distribution. Compared with the traditional threshold triggering method, this application can better handle extreme data and potential nonlinear relationships, thereby improving the ability to predict equipment or system failures. The SVM parameters are optimized through independent component analysis (ICA), and the model parameters are automatically adjusted to adapt to different substation equipment operating conditions, thereby improving the adaptability of the diagnostic method.

[0063] In some embodiments, monitoring data of the substation equipment is obtained, and an optimization target is determined to determine an objective function, wherein the optimization target satisfies the following relationship:

[0064]

[0065] Wherein, ω is the failure cycle statistic of the substation equipment, C is the factor of the substation equipment, ξ i is the fault degree index of the i-th substation equipment, is the fault impact of the i-th substation equipment.

[0066] Specifically, real-time monitoring data of the equipment is collected from the monitoring system (such as the SCADA system) of the substation, including but not limited to key performance indicators such as current, voltage, temperature, vibration, and pressure, to ensure that the collected data covers a sufficient time period to facilitate the analysis of the failure cycle statistics. According to the operating characteristics and failure history of the substation equipment, the optimization goal is determined, that is, minimizing the failure risk and maintenance cost of the equipment. According to the optimization goal, an objective function is constructed, which will be used to evaluate the operating status and failure risk of the equipment. The relationship of the objective function is shown in (1.2). The collected monitoring data is analyzed in time series to calculate the failure cycle of each statistic, that is, the frequency and regularity of equipment failures in a period of time. For example, the number of failures of the equipment in each month in the past year can be calculated to obtain the failure cycle statistics. According to the analysis of the failure data, a failure degree index fi is assigned to each statistic, which reflects the severity of the failure of the equipment in a certain aspect. The impact of each fault on the operation of the equipment and the stability of the power grid is evaluated to obtain the fault impact and calculate the objective function.

[0067] In some embodiments, after the optimization target is determined, the optimization target is dualized, and the dualized form satisfies the following relationship:

[0068]

[0069] Among them, Ψ(α) is the dual maximum function, y j is the index of the jth statistic, and l is the number of indicators of the statistic.

[0070] Specifically, after determining the optimization objective, it is transformed into a dual problem so that it can be solved using a dual algorithm. Dualization is usually used to simplify problems or improve solution efficiency, especially when dealing with optimization problems such as support vector machines (SVMs). For dualization, we construct a Lagrangian function, see (1.3). The partial derivatives of the Lagrangian function are calculated separately, and these partial derivatives are set to zero to obtain a set of equations. Substituting the above equations into the Lagrangian function, we obtain the expression of the dual problem. Use a dual algorithm (such as the sequential minimum optimization algorithm SMO) to solve the dual problem and find the value that maximizes Ψ(α). The dualization method can effectively transform the original optimization problem and make it easier to solve. Especially when dealing with large-scale data and high-dimensional problems, dualization can significantly improve computational efficiency and solution quality.

[0071] In some embodiments, S2, based on ICA, the step of performing SVM parameter optimization on the objective function specifically includes:

[0072] Randomly select some points on the function and initialize the empire;

[0073] the removal of colonies to relevant imperialist countries;

[0074] If there is a colony in the empire that costs less than the imperialist, swap the positions of the colony and the imperialist;

[0075] Calculate the total cost of all empires;

[0076] Take the weakest colony from the weakest empire and give it to the empire most likely to have him, and destroy the weak empire.

[0077] Specifically, a series of points are randomly selected within the definition domain of the objective function, which represent different SVM parameter combinations, and each point is used as an initial empire. Each empire is given a certain number of initial colonies. The number of colonies can be determined according to the initial strength of the empire. For each colony, its similarity or correlation with each imperialist country is calculated, and the colony is assigned to the most relevant imperialist country. Among them, this step can be achieved by comparing the SVM parameter differences between colonies and imperialism. The smaller the parameter difference, the higher the correlation. Within each empire, the cost of imperialism and all its colonies (i.e., the value of the objective function) is compared. If it is found that the cost of any colony is lower than that of the imperialist country to which it belongs, the positions of the two are exchanged, so that the parameter combination with lower cost becomes the new imperialism. For each empire, its total cost including imperialism and all colonies is calculated. -The total cost can be the average of the cost of imperialism and the cost of all colonies, or other forms of comprehensive cost assessment. From all empires, the empire with the highest total cost, i.e., the weakest empire, is identified. In this weakest empire, the colony with the highest cost, i.e., the weakest colony, is selected, and this weakest colony is assigned to the empire that is most likely to own it, which is usually the imperialist country closest to the colony. After the redistribution of colonies is completed, the weakest empire, that is, the empire with the highest total cost, is eliminated. This process simulates the process of imperial competition and annexation, and the ultimate goal is to leave the empire with the lowest cost and the best performance.

[0078] Through the above embodiments, the ICA algorithm can effectively optimize the SVM parameters and improve the accuracy of substation equipment fault diagnosis. This method can find the optimal or near-optimal SVM parameter combination in a complex parameter space by simulating the process of imperial competition and colonial allocation, thereby improving the performance of the fault diagnosis system.

[0079] In some embodiments, after S3, SVM parameters are optimized, the step of performing fault diagnosis on the substation equipment based on nonlinear SVM specifically includes:

[0080] Use PCA to perform linear transformation on the initial sample set, use the cumulative contribution rate after transformation to select the principal component, and perform dimensionality reduction processing on the sample set after linear transformation based on the selected principal component to obtain a new sample set;

[0081] Centrally training the new samples, wherein an objective function is formed by the new sample set and a multi-classification support vector machine with unknown parameters;

[0082] Use ICA to find the optimal solution of the objective function to determine the optimal parameters of SVM for smart substation fault classification;

[0083] Input the test sample into the support vector machine model and calculate its predicted value.

[0084] Specifically, PCA is used to preprocess the sample set. The initial sample set collected from the substation equipment contains monitoring data of multiple dimensions. Principal component analysis (PCA) is applied to perform linear transformation on the initial sample set to identify the main change direction in the data, calculate the cumulative contribution rate of each principal component after transformation, select the principal component whose cumulative contribution rate reaches a preset threshold (e.g., 95%), and perform dimensionality reduction processing on the sample set after linear transformation based on the selected principal component to obtain a set of new sample sets, which retains most of the information of the original data but has a lower dimension. Train the new sample set, use the new sample set after dimensionality reduction to train the SVM model, and construct the objective function of the multi-classification SVM, which contains unknown parameters that will determine the performance of the SVM model.

[0085] Use ICA to optimize SVM parameters and apply the Imperial Competition Algorithm (ICA) to find the optimal solution of the objective function. During the ICA process, multiple empires (i.e., different SVM parameter combinations) are initialized and allowed to compete and evolve in the parameter space. Through an iterative process, the ICA algorithm will find the optimal SVM parameter combination that maximizes or minimizes the objective function. After the SVM model training is completed, the test sample is input into the SVM model optimized by ICA. The SVM model calculates the predicted value of the test sample, i.e., the fault category, based on the optimized parameters. For example, a data set containing 1,000 samples was collected from substation equipment, each with 50 features. PCA was used to transform the data set, and it was found that the cumulative contribution rate of the first 10 principal components reached 95%. These 10 principal components were selected for dimensionality reduction to obtain a new sample set. On the new sample set, a multi-classification SVM objective function was constructed, and the ICA algorithm was used to optimize the SVM penalty parameter C and kernel function parameter γ. After ICA optimization, we obtained the optimal SVM parameter combination and input a test sample of unknown fault category into the optimized SVM model. The model output predicted the fault category as "mild overheating".

[0086] In some embodiments, when performing linear transformation on the initial sample set using PCA, the RBF kernel function is selected as the kernel function of the training model.

[0087] Specifically, when using principal component analysis (PCA) to perform linear transformation on the initial sample set, the radial basis function (RBF) kernel function is selected as the training model kernel function. For example, the initial sample set collected from the substation equipment contains monitoring data of multiple dimensions, such as temperature, pressure, current, etc. The initial sample set is linearly transformed using principal component analysis (PCA) to reduce the data dimension and extract the most important features. The covariance matrix of the sample set is calculated, and its eigenvalues ​​and eigenvectors are found. According to the size of the eigenvalues, the first k principal components are selected, which can explain most of the variability of the data (for example, the cumulative contribution rate reaches 95%).

[0088] On the new sample set after dimensionality reduction, the radial basis function (RBF) is selected as the kernel function of the SVM model. The RBF kernel function can map the input space to the high-dimensional feature space, making the originally linearly inseparable data separable in this space. The SVM model is trained using the new sample set after dimensionality reduction and the selected RBF kernel function. During the training process, the parameter γ of the RBF kernel function needs to be determined, which is usually done through methods such as cross-validation. At the same time, the penalty parameter C of the SVM needs to be determined, which controls the model's tolerance for misclassification.

[0089] The embodiment of the present application further provides a substation equipment fault diagnosis system based on SVM, and the substation equipment fault diagnosis system based on SVM includes:

[0090] The objective function determination module is configured to obtain monitoring data of the substation equipment, perform nonlinear transformation in combination with the SVM, and determine the objective function;

[0091] A parameter optimization module is configured to perform SVM parameter optimization on the objective function based on ICA;

[0092] The fault diagnosis module is configured to perform fault diagnosis on the substation equipment based on nonlinear SVM after SVM parameter optimization.

[0093] Specifically, the substation equipment fault diagnosis system based on SVM includes three main modules: objective function determination module, parameter optimization module and fault diagnosis module. The objective function determination module obtains real-time monitoring data from the substation monitoring system (such as SCADA system), including key performance indicators such as current, voltage, temperature, vibration, etc. The objective function determination module inputs these monitoring data into the preprocessing unit for data cleaning and normalization. The objective function determination module uses principal component analysis (PCA) to reduce the dimension of the data and extract the main features. Combined with the nonlinear transformation capability of SVM, the module constructs the objective function, which aims to minimize the classification error while maximizing the classification interval.

[0094] The parameter optimization module uses the Imperial Competition Algorithm (ICA) to optimize the SVM parameters of the objective function. A set of empires (i.e., parameter combinations) are randomly initialized in the parameter space of the objective function. Through an iterative process, the module moves colonies (variants of parameter combinations) to related imperialist countries (the current optimal parameter combination) and exchanges them according to the cost function to find the optimal parameters, calculate the total cost of all empires, and select the weakest colony from the weakest empire and give it to the empire most likely to own it, while eliminating the weak empires. The optimal parameters of the SVM, including the penalty parameter C and the kernel function parameter γ, are determined.

[0095] The fault diagnosis module uses the optimized SVM parameters to diagnose faults in substation equipment. The preprocessed new sample set is input into the trained nonlinear SVM model. The SVM model calculates the prediction value based on the input sample data and outputs the fault diagnosis result. If the diagnosis result shows that the equipment is faulty, the module will trigger an alarm and provide recommended maintenance measures.

[0096] In some embodiments, the SVM-based substation equipment fault diagnosis system includes:

[0097] The linear transformation module is configured to perform linear transformation on the initial sample set through PCA, select the principal component using the cumulative contribution rate after the transformation, and perform dimensionality reduction processing on the sample set after the linear transformation based on the selected principal component.

[0098] Specifically, the linear transformation module is a key component in the fault diagnosis system of substation equipment based on SVM. Its main function is to achieve data dimensionality reduction by performing principal component analysis (PCA) on the initial sample set. The real-time monitoring data of the equipment is collected from the monitoring system of the substation (such as the SCADA system) to form an initial sample set. These data may include indicators of multiple dimensions such as current, voltage, temperature, humidity, vibration, etc. The initial sample set is preprocessed, including data cleaning (removing outliers and missing values) and normalization (scaling the data to a fixed range, such as 0 to 1). The linear transformation module applies PCA to perform linear transformation on the preprocessed sample set. PCA transforms a set of possibly related variables into a set of linearly unrelated variables through orthogonal transformation. These new variables are called principal components. The module calculates the variance contribution rate of each principal component and arranges them in descending order. The module selects the principal components based on the cumulative contribution rate of the principal components. For example, the first few principal components with a cumulative contribution rate of 95% are selected. After calculation, the cumulative contribution rate of the first five principal components reaches 95%, then these principal components will be selected as the basis for dimensionality reduction.

[0099] Based on the selected principal components, the linear transformation module performs dimensionality reduction on the sample set. This means that the original data will be projected into a new feature space defined by these principal components, and the sample set after dimensionality reduction will contain fewer feature dimensions, but still retain most of the information of the original data.

[0100] The present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the substation equipment fault diagnosis method based on SVM as described in any one of the above items is implemented.

[0101] It should be noted that the storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a ROM (Read-Only Memory), a RAM (Random Access Memory), a magnetic disk or an optical disk.

[0102] It will be appreciated by those skilled in the art that all or some of the steps, systems, and functional modules / units in the methods disclosed above may be implemented as software, firmware, hardware, and appropriate combinations thereof. In a hardware implementation, the division between the functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed by several physical components in cooperation. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or may be implemented as hardware, or may be implemented as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable storage medium, which may include a computer-readable storage medium (or a non-transitory medium) and a communication medium (or a temporary medium).

[0103] As is known to those of ordinary skill in the art, the term computer-readable storage media includes volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules or other data). Computer-readable storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, it is known to those of ordinary skill in the art that communication media typically contain computer-readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transmission mechanism, and may include any information delivery medium.

[0104] Exemplarily, the computer-readable storage medium may be an internal storage unit of the electronic device of the aforementioned embodiment, such as a hard disk or memory of the electronic device. The computer-readable storage medium may also be an external storage device of the electronic device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device.

[0105] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit the same. Although the present application has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to replace some of the technical features therein by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions claimed to be protected by the present application.

Claims

1. A substation equipment fault diagnosis method based on SVM, characterized in that: The substation equipment fault diagnosis method based on SVM comprises the following steps: S1. Obtain monitoring data of substation equipment, perform nonlinear transformation in combination with SVM, and determine the objective function; S2. Optimizing SVM parameters of the objective function based on ICA; S3. After the SVM parameters are optimized, fault diagnosis of the substation equipment is performed based on nonlinear SVM; Among them, the objective function satisfies the following relationship: Among them, f(x) is the objective function, α i is the horizontal intercept of the ith statistic of the substation equipment, b is the vertical intercept of the statistic of the substation equipment, y i is the index of the i-th statistic, K(x,x i ) is the fault threshold function of the substation equipment, x i is the parameter of the i-th statistic.

2. The method for substation equipment fault diagnosis based on SVM according to claim 1 is characterized in that: Obtain monitoring data of substation equipment, determine the optimization target, and determine the objective function, where the optimization target satisfies the following relationship: Wherein, ω is the failure cycle statistic of the substation equipment, C is the factor of the substation equipment, ξ i is the fault degree index of the i-th substation equipment, is the fault impact of the i-th substation equipment.

3. The SVM-based substation equipment fault diagnosis method according to claim 2 is characterized in that: After the optimization target is determined, the optimization target is dualized, and the dualized form satisfies the following relationship: Among them, Ψ(α) is the dual maximum function, y j is the index of the jth statistic, and l is the number of indicators of the statistic.

4. The method for diagnosing substation equipment faults based on SVM according to claim 1, characterized in that: S2, based on ICA, the step of performing SVM parameter optimization on the objective function specifically comprises: Randomly select some points on the function and initialize the empire; the removal of colonies to relevant imperialist countries; If there is a colony in the empire that costs less than the imperialist, swap the positions of the colony and the imperialist; Calculate the total cost of all empires; Take the weakest colony from the weakest empire and give it to the empire most likely to have him, and destroy the weak empire.

5. The method for diagnosing substation equipment faults based on SVM according to claim 3 is characterized in that: S3. After the SVM parameters are optimized, the step of performing fault diagnosis on the substation equipment based on the nonlinear SVM specifically includes: Use PCA to perform linear transformation on the initial sample set, use the cumulative contribution rate after transformation to select the principal component, and perform dimensionality reduction processing on the sample set after linear transformation based on the selected principal component to obtain a new sample set; Centrally training the new samples, wherein an objective function is formed by the new sample set and a multi-classification support vector machine with unknown parameters; Use ICA to find the optimal solution of the objective function to determine the optimal parameters of SVM for smart substation fault classification; Input the test sample into the support vector machine model and calculate its predicted value.

6. The method for diagnosing substation equipment faults based on SVM according to claim 5 is characterized in that: When using PCA to perform linear transformation on the initial sample set, the RBF kernel function is selected as the kernel function of the training model.

7. A substation equipment fault diagnosis system based on SVM, characterized in that: The substation equipment fault diagnosis system based on SVM includes: The objective function determination module is configured to obtain monitoring data of the substation equipment, perform nonlinear transformation in combination with the SVM, and determine the objective function; A parameter optimization module is configured to perform SVM parameter optimization on the objective function based on ICA; The fault diagnosis module is configured to perform fault diagnosis on the substation equipment based on nonlinear SVM after SVM parameter optimization.

8. The SVM-based substation equipment fault diagnosis system according to claim 7 is characterized in that: The substation equipment fault diagnosis system based on SVM includes: The linear transformation module is configured to perform linear transformation on the initial sample set through PCA, select the principal component using the cumulative contribution rate after the transformation, and perform dimensionality reduction processing on the sample set after the linear transformation based on the selected principal component.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the SVM-based substation equipment fault diagnosis method according to any one of claims 1 to 7 is implemented.

Images