A multi-dimensional information event recognition method and system based on a multiple kernel support vector machine

Abstract

The application discloses a kind of multi-dimensional information event identification method and system based on multiple kernel support vector machine, the method of the present application includes extracting multi-dimensional feature vector from power grid, including part or all in time domain part characteristic value and frequency domain part characteristic value, time domain part characteristic value includes the effective value of voltage and current, skewness, kurtosis, frequency domain part characteristic value includes the amplitude spectrum of voltage and current, phase spectrum, the amplitude of specified hth harmonic;Multi-dimensional feature vector is input into multiple kernel support vector machine to realize multi-dimensional information event identification, the kernel function of multiple kernel support vector machine is obtained by the weighted fusion of the kernel function of Gaussian kernel and quantum kernel, and the quantum kernel is the square modulus of the inner product between the corresponding quantum states of different feature vectors in the multi-dimensional feature vector, which is used to quantify the global structural similarity of the feature vector.The present application aims to overcome the performance deficiency of traditional methods in event identification of power system under small sample condition, and improve the robustness and accuracy of event identification.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of power grid disturbance identification, and particularly relates to a multi-dimensional information event identification method and system based on a multiple kernel support vector machine. BACKGROUND

[0002] Inverter resources are increasingly occupying a higher proportion in power systems as new energy power generation systems join the grid. In this context, disturbances such as voltage fluctuations, frequency drops, and waveform distortions caused by inverters pose serious challenges to event detection and classification. It is worth noting that such disturbances caused by inverters are prone to trigger chain reactions, further threatening the stability of voltage and frequency in power systems. Therefore, it is crucial to identify event types and eliminate them in a timely manner. Neural network-related event identification methods require a large amount of data to complete event identification. For example, a multi-task time-frequency transformation network for identifying abnormal events such as power generation output, phase shift, and load reduction; a method that can integrate physical information and measurement data GNN, which identifies abnormalities by utilizing system impedance and network topology. However, these methods cannot meet the needs of event identification under small sample conditions. Traditional support vector machines perform well in small sample event identification, but they use a single kernel function, such as the Gaussian kernel, which has significant limitations, such as insufficient small sample generalization ability, insufficient utilization of multi-dimensional features, and low global structure sensitivity. Therefore, under small sample conditions, it is crucial to develop a method that meets the requirements of event identification accuracy and robustness. SUMMARY

[0003] The technical problem to be solved by the present application is to provide a multi-dimensional information event identification method and system based on a multiple kernel support vector machine to overcome the performance deficiencies of traditional methods in event identification under small sample conditions in power systems and improve the robustness and accuracy of event identification.

[0004] To solve the above technical problems, the technical scheme adopted by the present application is as follows:

[0005] A multi-dimensional information event identification method based on a multiple kernel support vector machine, comprising the following steps:

[0006] S101, extracting a multi-dimensional feature vector from the power grid, including part or all of time domain part feature values and frequency domain part feature values, the time domain part feature values including part or all of effective values, skewness, and kurtosis of voltage and current, and the frequency domain part feature values including part or all of amplitude spectrum, phase spectrum, and specified hth harmonic amplitude of voltage and current;

[0007] S102, input the multi-dimensional feature vector into a multiple kernel support vector machine to realize multi-dimensional information event recognition, a kernel function of the multiple kernel support vector machine being obtained by weighted fusion of kernel functions of a Gaussian kernel and a quantum kernel, the quantum kernel being a square modulus of an inner product between corresponding quantum states of different feature vectors in the multi-dimensional feature vector for quantifying global structural similarity of the feature vectors.

[0008] Optionally, a function expression of the kernel function of the quantum kernel is:

[0009] ,

[0010] is the kernel function of the quantum kernel, and are the i-th dimensional feature vector and the j-th dimensional feature vector in the multi-dimensional feature vector, respectively, denotes an inner product between a quantum state of the feature vector and of the feature vector , is a probability that all quantum bits are measured as , is a conjugate of a quantum feature map of the feature vector , is a quantum feature map of the feature vector , is an initial state of quantum bits.

[0011] Optionally, a function expression of the kernel function of the multiple kernel support vector machine is:

[0012] ,

[0013] ,

[0014] wherein, is the kernel function of the multiple kernel support vector machine, and are kernel functions of the Gaussian kernel for processing time domain features and frequency domain features, and are kernel functions of the quantum kernel for processing time domain features and frequency domain features, , , , , , is a weight coefficient.

[0015] ​Optionally, the function expression for inputting the multidimensional feature vector into a multi-kernel support vector machine to achieve multidimensional information event recognition is:

[0016] ,

[0017] in, This represents the multidimensional information event recognition results from a multi-kernel support vector machine. For symbolic functions, For the weight vector, the superscript... For transpose operation, For the kernel function of a multi-kernel support vector machine, This is a bias term.

[0018] Optionally, the training of the multi-kernel support vector machine includes:

[0019] S201, extract multidimensional feature vectors and their labels from multidimensional information event samples from the power grid to construct a training dataset;

[0020] S202, Adjust weighting coefficients , , , , , And construct the kernel function for the multi-kernel support vector machine;

[0021] S203, train a multi-kernel support vector machine using the training dataset and optimize the model parameters according to the following formula:

[0022] ,

[0023] ,

[0024] ,

[0025] in, For the weight vector, For bias terms, is the penalty coefficient in the kernel function of the Gaussian kernel. The number of samples in the training dataset, For the first Slack variables for each sample, For the first The label of each sample The kernel function is the kernel function of the multi-kernel support vector machine; the model parameters include the weight coefficients in the kernel function of the multi-kernel support vector machine. , , , , , and a bias term in a kernel function of a Gaussian kernel and a penalty coefficient

[0026] S204, it is judged whether a preset ending condition is reached. If the preset ending condition is not reached, the step S202 is jumped to continue the training. Otherwise, the training is ended and exited.

[0027] Optionally, a function expression of the preset ending condition in the step S204 is:

[0028]

[0029] wherein, e is a prediction error, is a new model parameter, is a current model parameter, and a multiple kernel support vector machine.

[0030] Optionally, a function expression of the kernel function of the Gaussian kernel is:

[0031]

[0032] wherein, K(x, x') is the kernel function of the Gaussian kernel, and are a m-dimensional feature vector and a n-dimensional feature vector, respectively, is a Gaussian kernel width. In addition, the embodiment also provides a multiple kernel support vector machine based multi-dimensional information event recognition system, which comprises a microprocessor and a memory connected with each other. The microprocessor is programmed or configured to execute the multiple kernel support vector machine based multi-dimensional information event recognition method.

[0033] In addition, the embodiment also provides a computer readable storage medium, which stores a computer program or instructions. The computer program or instructions are programmed or configured to execute the multiple kernel support vector machine based multi-dimensional information event recognition method by a processor.

[0034] In addition, the embodiment also provides a computer program product, which comprises a computer program or instructions. The computer program or instructions are programmed or configured to execute the multiple kernel support vector machine based multi-dimensional information event recognition method by a processor.

[0035] In addition, the embodiment also provides a computer program product, which comprises a computer program or instructions. The computer program or instructions are programmed or configured to execute the multiple kernel support vector machine based multi-dimensional information event recognition method by a processor.

[0036] ​​​​​​Compared with the prior art, the method has the following beneficial effects: the method comprises extracting a multi-dimensional feature vector from a power grid, the multi-dimensional feature vector comprising part or all of time domain partial feature values and frequency domain partial feature values, the time domain partial feature values comprising effective values, skewness and kurtosis of voltage and current, and the frequency domain partial feature values comprising amplitude spectrum, phase spectrum and specified hth harmonic amplitude of voltage and current; and the multi-dimensional feature vector is input into a multiple kernel support vector machine to realize multi-dimensional information event identification, the kernel function of the multiple kernel support vector machine being obtained by weighted fusion of kernel functions of a Gaussian kernel and a quantum kernel, and the quantum kernel being a square modulus of an inner product between corresponding quantum states of different feature vectors in the multi-dimensional feature vector for quantifying global structural similarity of the feature vector. Compared with a traditional classification method, the method has the advantages of being able to use multiple kernel functions to map each input feature, thereby improving classification accuracy, having high accuracy under a small sample condition, effectively overcoming insufficient event identification performance of a traditional method under a small sample condition of a power system, and improving robustness and accuracy of event identification. BRIEF DESCRIPTION OF DRAWINGS

[0037] Figure 1 FIG. 1 is a schematic diagram of a training process of a multiple kernel support vector machine in an embodiment of the present application.

[0038] Figure 2 FIG. 4 is a schematic diagram of performance under different training data proportions in an embodiment of the present application.

[0039] Figure 3 FIG. 5 is a schematic diagram of performance comparison of different event identification methods in an embodiment of the present application. DETAILED DESCRIPTION

[0040] In order to enable personnel in the technical field to better understand the technical solutions of the present application, the technical solutions of the present application will be further described in detail below with reference to the accompanying drawings of the embodiments of the present application.

[0041] As shown in FIG. 1, the multi-dimensional information event identification method based on the multiple kernel support vector machine in the embodiment comprises the following steps: Figure 1

[0042] S101, extracting a multi-dimensional feature vector from a power grid, the multi-dimensional feature vector comprising part or all of time domain partial feature values and frequency domain partial feature values, the time domain partial feature values comprising part or all of effective values, skewness and kurtosis of voltage and current, and the frequency domain partial feature values comprising part or all of amplitude spectrum, phase spectrum and specified hth harmonic amplitude of voltage and current;

[0043] ​S102, input the multidimensional feature vector into the multi-kernel support vector machine (QMSVM) to realize multidimensional information event recognition. The kernel function of the multi-kernel support vector machine is obtained by weighted fusion of the kernel functions of Gaussian kernel and quantum kernel. The quantum kernel is the square modulus of the inner product between the corresponding quantum states of different feature vectors in the multidimensional feature vector to quantize the global structural similarity of the feature vector.

[0044] In this embodiment, the kernel function of the Gaussian kernel is expressed as follows:

[0045] ,

[0046] in, Let be the kernel function of the Gaussian kernel. and These are multidimensional feature vectors. The first in 3D eigenvectors and the 1st eigenvector 3D feature vectors The width of the Gaussian kernel. , used to control the mapping results.

[0047] The intrinsic properties of kernel functions significantly impact recognition performance. Gaussian kernels primarily rely on the Euclidean distance between samples, exhibiting strong local learning capabilities. However, their sensitivity to the global structure of the data is limited. Therefore, using only Gaussian kernels for event recognition based on multidimensional information of event samples can result in relatively large errors. Introducing quantum kernels to increase sensitivity to the global data can improve the accuracy of event recognition. This involves transforming the dimensional feature vector... The method for converting to quantum states for computation is as follows: convert the multidimensional eigenvectors Mapping to quantum state ,in The quantum superposition representation of the features is achieved by performing quantum gate operations on the initial all-zero state through a quantum feature map (quantum circuit):

[0048] ,

[0049] in, Indicates an initial all-zero state. For feature vectors The quantum feature map. With the input being... and In this case, the quantum kernel is defined as the square modulus of the inner product between its corresponding quantum states, quantizing the global structural similarity of the eigenvectors. In this embodiment, the kernel function of the quantum kernel is expressed as:

[0050] ,

[0051] in, is the kernel function of the quantum kernel, and are the i-th and j-th eigenvectors in the multi-dimensional eigenvector , denotes the inner product between the quantum state and the quantum state , and , is the probability that all the qubits are measured as , is the conjugate of the quantum eigenvector , is the quantum eigenvector , is the initial state of the qubits; is the tensor product symbol; denotes the n-fold tensor product operation on , i.e., the combination of n states. The quantum kernel can capture high-order correlations between features by utilizing the quantum entanglement property, making up for the lack of global structure perception of the Gaussian kernel. Based on the above, the Gaussian kernel and the quantum kernel are combined into a multi-kernel. It should be noted that the quantum eigenvector (quantum circuit) is a publicly known method, and details can be found in Gong Jing, He Min, Yao Zeqing. Quantum circuit implementation of entanglement exchange [J]. Communication technology, 2008, 41(6): 3. DOI:10.3969 / j.issn.1002-0802.2008.06.027.

[0052] The weighted fusion of the kernel functions of the Gaussian kernel and the quantum kernel into a multi-kernel can be performed in the required fusion manner. For example, as an optional embodiment, the function expression of the kernel function of the multi-kernel support vector machine in the present embodiment is:

[0053] ,

[0054] ,

[0055] wherein, is the kernel function of the multi-kernel support vector machine, and are the kernel functions of the Gaussian kernel for processing the time domain features and the frequency domain features, respectively, and are the kernel functions of the quantum kernel for processing the time domain features and the frequency domain features, respectively, , , , , , are weight coefficients. Wherein, is the weight of the time domain feature, is the weight of the frequency domain feature, and are the weights of the Gaussian kernel part in the time domain feature and the frequency domain feature respectively, and are the weights of the quantum kernel part in the time domain feature and the frequency domain feature respectively. New events are highly nonlinear due to environmental influences and change rapidly over time. At the microscopic level, quantum effects in power electronic devices can affect the dynamic behavior of the system, and these features have complex quantum correlations in high-dimensional space, and the traditional kernel function has insufficient feature expression capability. The kernel function of the multiple kernel support vector machine in the embodiment can realize differentiated modeling of different dimensional features by adaptively adjusting the weights, and improve the expression capability of mixed features.

[0056] In the embodiment, the function expression of inputting the multi-dimensional feature vector into the multiple kernel support vector machine to realize multi-dimensional information event recognition is:

[0057] ,

[0058] Wherein, is the multi-dimensional information event recognition result of the multiple kernel support vector machine, is a sign function, is a weight vector, and the superscript is a transpose operation, is a kernel function of the multiple kernel support vector machine, is a bias term.

[0059] As shown in Figure 1 , the training of the multiple kernel support vector machine in the embodiment includes:

[0060] S201, constructing a training data set from the multi-dimensional feature vector and the label of the multi-dimensional information event sample extracted from the power grid, and the multi-dimensional information event sample can be expressed as:

[0061] ,

[0062] Wherein, are the first dimensional feature vector and the label in the multi-dimensional feature vector , is the number of dimensions;

[0063] S202, adjusting the weight coefficients , , , , , And construct the kernel function for the multi-kernel support vector machine (QMSVM);

[0064] S203, In order to learn the decision plane for classification, a multi-kernel support vector machine (QMSVM) is trained using the training dataset, and the model parameters are optimized according to the following formula:

[0065] ,

[0066] ,

[0067] ,

[0068] in, For the weight vector, For bias terms, is the penalty coefficient in the kernel function of the Gaussian kernel (used to balance marginal maximization and classification error). The number of samples in the training dataset, For the first Slack variables for each sample, For the first The label of each sample The kernel function of the multi-kernel support vector machine is given above. By introducing the dual Lagrangian function, the model parameters of the proposed QMSVM can be obtained through partial derivatives. These model parameters include the weight coefficients in the kernel function of the multi-kernel support vector machine. , , , , , And the bias term in the kernel function of the Gaussian kernel. and penalty coefficient ;

[0069] S204, determine whether the preset termination condition has been met. If the preset termination condition has not been met, jump to step S202 to continue training; otherwise, end and exit.

[0070] like Figure 1 As shown, the function expression for the preset termination condition in step S204 is:

[0071] ,

[0072] in, For prediction error, , For the new model parameters, For the current model parameters, represents a multiple kernel support vector machine. The training of the multiple kernel support vector machine can use a simulated annealing algorithm to optimize the parameters to improve the identification performance. Through the above preset end condition, the "temperature" parameter can be gradually reduced, and the search range can be narrowed until the global optimal parameters are converged.

[0073] To verify the multiple kernel support vector machine of the embodiment, the embodiment establishes a distributed simulation model containing 14 photovoltaics based on the IEEE123 bus test feeder in OpenDSS and ATP-EMTP. The load conditions, photovoltaic penetration level and node position are specially configured to simulate the measurement uncertainty commonly seen in actual power systems. In addition, in order to test the robustness of the multiple kernel support vector machine of the embodiment under various operating conditions, the load level, fault location and switch stage are diversified during simulation, such as simulating typical interference conditions such as three-phase, single-phase and line-to-line faults. In order to verify the robustness of the multiple kernel support vector machine (QMSVM) proposed in the embodiment, the performance of the existing Sigmoid kernel SVM (SSVM), Gaussian kernel SVM (GSVM), Quantum kernel SVM (QSVM) and the multiple kernel support vector machine proposed in the embodiment under different training data proportions is compared, which means that different number of training samples are used in the training process. The final detection results are as shown in Figure 2 Figure 2 ​As can be seen, across all training data proportions, the classification accuracy of the proposed multi-kernel support vector machine consistently outperforms other support vector machine methods. Notably, the accuracy of Gaussian kernel SVM and Quantum kernel SVM is slightly lower than that of the proposed multi-kernel support vector machine. This indicates that both global and local information contribute to anomaly detection. The key difference lies in the careful adjustment of the weight coefficients to balance the impact of the two tasks. Furthermore, as the training sample proportion increases, the accuracy of all methods improves accordingly, with the proposed multi-kernel support vector machine consistently achieving the highest accuracy. Specifically, at training ratios of 40% and 70%, the accuracy of the proposed multi-kernel support vector machine reaches 92.44% and 93.97%, respectively. This demonstrates that the proposed multi-kernel support vector machine exhibits good stability and robustness even with small sample sizes. To further verify the performance of the multi-kernel support vector machine proposed in this embodiment, five mainstream machine learning methods were selected for comparison: decision tree (DT), Naive Bayes (NB), K-nearest neighbor (KNN), gradient boosting decision tree (GBDT), and support vector machine (SVM). These methods were used to identify multi-dimensional event types, and the performance of each method was comprehensively evaluated using four metrics: precision, recall, and F1 score. The comparison results are as follows: Figure 3 As shown. From Figure 3 As can be seen, the multi-kernel support vector machine proposed in this embodiment outperforms other machine learning models on all evaluation metrics. Specifically, the multi-kernel support vector machine proposed in this embodiment achieves over 94% in precision, recall, and F1 score, significantly surpassing gradient boosting decision trees, Naive Bayes, decision trees, and K-nearest neighbors. Furthermore, compared to traditional support vector machines, the multi-kernel support vector machine proposed in this embodiment exhibits higher precision and F1 score. These results demonstrate the superior performance of the multi-kernel support vector machine proposed in this embodiment in event type detection.

[0074] In summary, the method of the embodiment includes extracting a multi-dimensional feature vector from the power grid, including some or all of the time domain partial feature values and the frequency domain partial feature values, the time domain partial feature values including the effective values, skewness, and kurtosis of the voltage and current, and the frequency domain partial feature values including the amplitude spectrum, phase spectrum, and specified hth harmonic amplitude of the voltage and current; inputting the multi-dimensional feature vector into a multiple kernel support vector machine to realize multi-dimensional information event recognition, the kernel function of the multiple kernel support vector machine being obtained by weighted fusion of the kernel functions of a Gaussian kernel and a quantum kernel, the quantum kernel being the square modulus of the inner product between corresponding quantum states of different feature vectors in the multi-dimensional feature vector for quantifying the global structural similarity of the feature vectors. Compared with traditional classification methods, the method of the embodiment has the advantages of being able to use multiple kernel functions to map each input feature, thereby improving the classification accuracy, having high accuracy under small sample conditions, eliminating the time-consuming of large sample training and the difficulty of sample collection, effectively overcoming the insufficient event recognition performance of traditional methods for power systems under small sample conditions, and improving the robustness and accuracy of event recognition.

[0075] In addition, the embodiment also provides a multi-dimensional information event recognition system based on a multiple kernel support vector machine, including a microprocessor and a memory connected to each other, the microprocessor being programmed or configured to execute the multi-dimensional information event recognition method based on the multiple kernel support vector machine.

[0076] In addition, the embodiment also provides a computer readable storage medium having a computer program or instructions stored therein, the computer program or instructions being programmed or configured to execute the multi-dimensional information event recognition method based on the multiple kernel support vector machine by a processor.

[0077] In addition, the embodiment also provides a computer program product including a computer program or instructions, the computer program or instructions being programmed or configured to execute the multi-dimensional information event recognition method based on the multiple kernel support vector machine by a processor.

[0078] Those skilled in the art will appreciate that the technology provided herein is not limited to any particular form of implementation. The technology provided herein can be implemented in hardware, software, or a combination thereof. Those skilled in the art will appreciate that the technology provided herein can be implemented in a number of different embodiments, including method embodiments, system embodiments, and computer program product embodiments. The technology provided herein can be implemented in any combination of hardware, software, or a combination thereof. The technology provided herein can be implemented in a number of different ways, including as a computer program product stored on a computer readable storage medium, as a system on chips (SOCs), as an application specific integrated circuit (ASIC), or as a combination of the above. The technology provided herein can be implemented using any suitable hardware, software, firmware, or combination thereof. The technology provided herein can be implemented in one or more computer programs or one or more articles of manufacture that contain computer readable program code. The technology provided herein can be implemented using any suitable computer readable storage medium, including storage devices that are external or internal to a computer. Suitable computer readable storage mediums can include, but are not limited to, volatile memory, non-volatile memory, removable storage, and non-removable storage. Suitable computer readable storage mediums can include, but are not limited to, RAM, ROM, EEPROM, flash memory, or any other memory technology. Suitable computer readable storage mediums can include, but are not limited to, magnetic cassettes, magnetic tapes, magnetic disks, memory cards or sticks, optical storage media, or any other storage medium suitable for storing computer readable program code. The computer readable program code can be executed using any suitable computer processor, including a general purpose computer, a special purpose computer, an embedded computer, or any other computer. The computer readable program code can be executed using any suitable operating system, including a UNIX operating system, a LINUX operating system, a WINDOWS operating system, a MAC OS operating system, or any other operating system. The computer readable program code can be executed using any suitable computer programming language, including a high level programming language, a low level programming language, an object oriented programming language, a visual programming language, or any other computer programming language. Figure 1 one or more functions specified in the flow or flows and / or blocks Figure 1 one or more functions specified in the flow or flows and / or blocks Figure 1 one or more functions specified in the flow or flows and / or blocks Figure 1 one or more functions specified in the flow or flows and / or blocks Figure 1 one or more functions specified in the flow or flows and / or blocks Figure 1 one or more functions specified in the flow or flows and / or blocks

[0079] The above description is only preferred embodiments of the present application. The protection scope of the present application is not limited to the above-mentioned embodiments. Any technical scheme falling within the concept of the present application is within the protection scope of the present application. It should be noted that some improvements and refinements made by those skilled in the art without departing from the principle of the present application are also considered to be within the protection scope of the present application.

Claims

1. A multi-dimensional information event recognition method based on multiple kernel support vector machines, characterized in that, The method comprises the following steps: S101, extracting a multi-dimensional feature vector from the power grid, the multi-dimensional feature vector comprising time domain partial feature values and frequency domain partial feature values, the time domain partial feature values comprising some or all of effective values, skewness and kurtosis of voltage and current, and the frequency domain partial feature values comprising some or all of amplitude spectrum, phase spectrum and specified hth harmonic amplitude of voltage and current; S102, inputting the multi-dimensional feature vector into a multiple kernel support vector machine to realize multi-dimensional information event recognition, a kernel function of the multiple kernel support vector machine being obtained by weighted fusion of kernel functions of a Gaussian kernel and a quantum kernel, the quantum kernel being a square modulus of an inner product between corresponding quantum states of different feature vectors in the multi-dimensional feature vector for quantifying global structural similarity of the feature vectors; a function expression of the kernel function of the multiple kernel support vector machine being: ， ， wherein, is a kernel function of a multiple kernel support vector machine, are kernel functions of a Gaussian kernel processing time domain features and frequency domain features, respectively, are kernel functions of a quantum kernel processing time domain features and frequency domain features, respectively, are weight coefficients.​​​​​​​ 2.The multi-dimensional information event recognition method based on multiple kernel support vector machine according to claim 1, characterized in that, a function expression of the kernel function of the quantum kernel being: ， wherein, is a kernel function for a quantum nucleus, and are, respectively, an i-th dimensional eigenvector and a j-th dimensional eigenvector in a multi-dimensional eigenvector , denotes an inner product between a quantum state of an eigenvector and a quantum state , is a probability that all quantum bits are measured as , is a conjugate of a quantum eigendefinition of an eigenvector , is a quantum eigendefinition of an eigenvector , is an initial state of quantum bits. 3.The multi-dimensional information event recognition method based on multiple kernel support vector machine according to claim 1, characterized in that, a function expression of the inputting the multi-dimensional feature vector into the multiple kernel support vector machine to realize the multi-dimensional information event recognition being: ， wherein is a multi-dimensional information event recognition result of a multiple kernel support vector machine, is a sign function, is a weight vector, the superscript is a transposition operation, is a kernel function of a multiple kernel support vector machine, is a bias term. 4.The multi-dimensional information event recognition method based on multiple kernel support vector machine according to claim 1, wherein, training of the multiple kernel support vector machine comprising: S201, constructing a training data set by extracting a multi-dimensional feature vector of a multi-dimensional information event sample and a label thereof from the power grid; S202, adjusting the weight coefficient 、 、 、 、 、 and construct the kernel function of the multiple kernel support vector machine; S203, training the multiple kernel support vector machine by using the training data set and optimizing model parameters according to the following formula: ， ， ， in, For the weight vector, For bias terms, is the penalty coefficient in the kernel function of the Gaussian kernel. The number of samples in the training dataset, For the first Slack variables for each sample, For the first The label of each sample The kernel function is the kernel function of the multi-kernel support vector machine; the model parameters include the weight coefficients in the kernel function of the multi-kernel support vector machine. , , , , , And the bias term in the kernel function of the Gaussian kernel. and penalty coefficient ; S204, judging whether a preset ending condition is reached, if the preset ending condition is not reached, jumping to step S202 to continue training, otherwise ending and exiting.

5. The multi-dimensional information event recognition method based on multiple kernel support vector machines according to claim 4, characterized in that, a function expression of the preset ending condition in step S204 being: ， wherein, is the prediction error, , is the new model parameter, is the current model parameter, denotes a multiple kernel support vector machine. 6.The multi-dimensional information event recognition method based on multiple kernel support vector machine according to claim 1, wherein, a function expression of the kernel function of the Gaussian kernel being: ， wherein, is a kernel function of a Gaussian kernel, and are the i-th and j-th multi-dimensional feature vectors, respectively, is the i-th and j-th feature vector in the i-th and j-th multi-dimensional feature vectors, respectively, is a Gaussian kernel width.

7. A multi-dimensional information event recognition system based on multiple kernel support vector machines, comprising a microprocessor and a memory connected to each other, characterized in that, the microprocessor being programmed or configured to execute the multi-dimensional information event recognition method based on the multiple kernel support vector machine according to any one of claims 1-6.

8. A computer-readable storage medium having stored therein a computer program or instructions, characterized in that, the computer program or instructions being programmed or configured to execute the multi-dimensional information event recognition method based on the multiple kernel support vector machine according to any one of claims 1-6 by the processor.

9. A computer program product comprising computer programs or instructions, characterized in that, the computer program or instructions being programmed or configured to execute the multi-dimensional information event recognition method based on the multiple kernel support vector machine according to any one of claims 1-6 by the processor.

Images