Artificial Intelligence-Based Intelligent Power Grid Operation and Maintenance Monitoring Method and System

Abstract

The intelligent power grid operation and maintenance monitoring method and system based on artificial intelligence provided by the present invention relate to the technical field of data processing. In the present invention, data extraction processing can be first performed on a target power grid component to obtain historical power grid operation data corresponding to the target power grid component; then, the data correlation between the historical power grid operation data and each reference power grid operation data included in a pre-configured reference power grid operation data set is calculated respectively; finally, based on the data correlation between the historical power grid operation data and each reference power grid operation data, and in combination with the abnormality degree reference value pre-configured for each reference power grid operation data, an abnormality degree characterization value corresponding to the historical power grid operation data is analyzed and output, and then the target operation safety degree corresponding to the target power grid component is determined based on the abnormality degree characterization value. Based on the above content, the reliability of operation safety analysis can be improved to a certain extent.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular, to an intelligent power grid operation and maintenance monitoring method and system based on artificial intelligence. Background Art

[0002] Artificial Intelligence (AI) is to use digital computers or computing simulations controlled by digital computers to simulate, extend, and expand human intelligence, a theory, method, technology, and application system that perceives the environment, acquires knowledge, and uses knowledge to obtain the best results. In other words, artificial intelligence is a comprehensive technology in computer science that attempts to understand the essence of intelligence and produce a new intelligent machine that can respond in a way similar to human intelligence.

[0003] Machine Learning (ML) is an interdisciplinary subject that involves multiple disciplines such as probability theory, statistics, approximation theory, convex analysis, and algorithm complexity theory. It specifically studies how computers simulate or implement human learning behaviors to acquire new knowledge or skills and reorganize the existing knowledge structure to continuously improve their own performance. Machine learning is the core of artificial intelligence and the fundamental way to make computers intelligent, and its applications cover all fields of artificial intelligence. Machine learning and deep learning usually include technologies such as artificial neural networks, belief networks, reinforcement learning, transfer learning, inductive learning, and rote learning.

[0004] The operation and maintenance of the power grid generally involve many levels, all of which affect the orderly operation of the power grid. Therefore, reliable operation and maintenance of the power grid are required. Among them, on the basis of reliable operation and maintenance of the power grid, effective monitoring of the operation safety of the power grid is carried out so that when an abnormality occurs, maintenance can be carried out in a timely manner to avoid the abnormality evolving into a fault. However, in the prior art, generally, the operation data of a single power device is analyzed, such as threshold comparison, etc. In this way, it is easy to have a problem that the reliability of safety analysis is not high due to the relatively single basis for data analysis. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to provide an intelligent power grid operation and maintenance monitoring method and system based on artificial intelligence to improve the reliability of operation safety analysis to a certain extent.

[0006] To achieve the above purpose, the embodiments of the present invention adopt the following technical solutions:

[0007] An intelligent power grid operation and maintenance monitoring method based on artificial intelligence, comprising:

[0008] Extract data from the target power grid component to obtain the historical power grid operation data corresponding to the target power grid component. The historical power grid operation data includes the historical power parameters of each target power device in the target power grid component at multiple historical time points. The target power grid component includes multiple target power devices, and there is a correlation between the multiple target power devices;

[0009] Calculate the data correlation degree between the historical power grid operation data and each reference power grid operation data included in the pre-configured reference power grid operation data set respectively;

[0010] Based on the data correlation degree between the historical power grid operation data and each reference power grid operation data, and in combination with the abnormal degree reference value pre-configured for each reference power grid operation data, analyze and output the abnormal degree characterization value corresponding to the historical power grid operation data, and then determine the target operation safety degree corresponding to the target power grid component based on the abnormal degree characterization value.

[0011] In some preferred embodiments, in the above intelligent power grid operation and maintenance monitoring method based on artificial intelligence, the step of extracting data from the target power grid component to obtain the historical power grid operation data corresponding to the target power grid component includes:

[0012] Extract data from the target power grid component according to the historical time period to obtain the initial historical power grid operation data corresponding to the target power grid component. The initial historical power grid operation data includes the data of each time point in the historical time period of the target power grid component;

[0013] Sample the historical time period to form multiple historical time points. The historical time interval lengths between the multiple historical time points have a decreasing corresponding relationship along the time trend;

[0014] For each historical time point, extract the historical power parameters of each target power device in the target power grid component from the initial historical power grid operation data respectively to obtain the historical power grid operation data corresponding to the target power grid component.

[0015] In some preferred embodiments, in the above intelligent power grid operation and maintenance monitoring method based on artificial intelligence, the step of calculating the data correlation degree between the historical power grid operation data and each reference power grid operation data included in the pre-configured reference power grid operation data set respectively includes:

[0016] Extract a reference power grid operation data from multiple reference power grid operation data included in a pre-configured set of reference power grid operation data, and then use the reference power grid operation data and the historical power grid operation data as a data combination for calculating the relevance of data;

[0017] Load the historical power grid operation data and the reference power grid operation data included in the data combination into the optimized data mining neural network respectively, and use the optimized data mining neural network to respectively mine the key information mining results of the power grid operation data corresponding to the historical power grid operation data and the key information mining results of the power grid operation data corresponding to the reference power grid operation data;

[0018] Based on the key information mining results of the power grid operation data corresponding to the historical power grid operation data and the key information mining results of the power grid operation data corresponding to the reference power grid operation data, calculate and output the matching degree of the key information mining results corresponding to the data combination, as the data relevance between the historical power grid operation data and the reference power grid operation data included in the data combination.

[0019] In some preferred embodiments, in the above intelligent power grid operation and maintenance monitoring method based on artificial intelligence, the optimization process of the optimized data mining neural network includes:

[0020] Extract multiple typical power grid operation data, and extract the power grid operation data identification information corresponding to the multiple typical power grid operation data;

[0021] Based on the key information mining results of the power grid operation data corresponding to at least one typical power grid operation data and the corresponding power grid operation data identification information, mine the extended key information mining results of the power grid operation data corresponding to the at least one typical power grid operation data;

[0022] Optimize the pre-built initial data mining neural network according to the multiple typical power grid operation data and the extended key information mining results of the power grid operation data corresponding to the at least one typical power grid operation data, so as to form the corresponding optimized data mining neural network;

[0023] Among them, the step of extracting multiple typical power grid operation data and extracting the power grid operation data identification information corresponding to the multiple typical power grid operation data includes:

[0024] Extract multiple typical power grid operation data, and screen at least one first typical power grid operation data from the multiple typical power grid operation data, where the at least one first typical power grid operation data belongs to the typical power grid operation data that does not have actual power grid operation data identification information among the multiple typical power grid operation data; and, based on the power grid operation data key information mining results corresponding to the at least one first typical power grid operation data, classify and form corresponding at least one classification identification information; and, for each of the first typical power grid operation data, determine the power grid operation data identification information corresponding to the first typical power grid operation data according to the classification identification information corresponding to the first typical power grid operation data.

[0025] In some preferred embodiments, in the above-mentioned intelligent power grid operation and maintenance monitoring method based on artificial intelligence, the step of mining the extended power grid operation data key information mining results corresponding to the at least one typical power grid operation data based on the power grid operation data key information mining results corresponding to the at least one typical power grid operation data and the corresponding power grid operation data identification information includes:

[0026] Based on the power grid operation data key information mining results corresponding to the at least one typical power grid operation data and the corresponding power grid operation data identification information, mine the first power grid operation data key information mining results corresponding to the at least one typical power grid operation data;

[0027] Based on the first power grid operation data key information mining results corresponding to the at least one typical power grid operation data, analyze and output the second power grid operation data key information mining results corresponding to the at least one typical power grid operation data;

[0028] Based on the power grid operation data key information mining results corresponding to the at least one typical power grid operation data, the corresponding first power grid operation data key information mining results and the corresponding second power grid operation data key information mining results, determine the extended power grid operation data key information mining results corresponding to the at least one typical power grid operation data;

[0029] Among them, the step of mining the first power grid operation data key information mining results corresponding to the at least one typical power grid operation data based on the power grid operation data key information mining results corresponding to the at least one typical power grid operation data and the corresponding power grid operation data identification information includes:

[0030] Based on the grid operation data identification information corresponding to at least one typical grid operation data, the typical grid operation data with the same corresponding grid operation data identification information is attributed to a grid operation data set to form at least one grid operation data set; based on the mining results of the grid operation data key information corresponding to the typical grid operation data included in each of the grid operation data sets, the mining results of the grid operation data set key information corresponding to each of the grid operation data sets are mined; the mining results of the grid operation data set key information corresponding to the grid operation data sets corresponding to the at least one typical grid operation data are marked as the first grid operation data key information mining results of the at least one typical grid operation data pair.

[0031] In some preferred embodiments, in the above-mentioned intelligent grid operation and maintenance monitoring method based on artificial intelligence, the step of analyzing and outputting the second grid operation data key information mining results corresponding to the at least one typical grid operation data based on the first grid operation data key information mining results corresponding to the at least one typical grid operation data includes:

[0032] Filter out the to-be-confirmed grid operation data sets corresponding to the at least one typical grid operation data; and, based on the first grid operation data key information mining results corresponding to the at least one typical grid operation data and the mining results of the grid operation data set key information corresponding to the to-be-confirmed grid operation data sets, analyze the confirmed grid operation data sets corresponding to the at least one typical grid operation data in the to-be-confirmed grid operation data sets;

[0033] The mining results of the grid operation data set key information corresponding to the confirmed grid operation data sets corresponding to the at least one typical grid operation data are marked as the second grid operation data key information mining results corresponding to the at least one typical grid operation data.

[0034] In some preferred embodiments, in the above-mentioned intelligent grid operation and maintenance monitoring method based on artificial intelligence, the step of determining the extended grid operation data key information mining results corresponding to the at least one typical grid operation data based on the mining results of the grid operation data key information corresponding to the at least one typical grid operation data, the corresponding first grid operation data key information mining results, and the corresponding second grid operation data key information mining results includes:

[0035] Determine a target variable, and there is a positive correlation between the current value of the target variable and the optimization progress of optimizing the initial data mining neural network;

[0036] Based on the key information mining results of the grid operation data corresponding to the at least one typical grid operation data, the corresponding first key information mining results of the grid operation data, the corresponding second key information mining results of the grid operation data, and the target variable, determine the key information mining results of the extended grid operation data corresponding to the at least one typical grid operation data.

[0037] In some preferred embodiments, in the above intelligent grid operation and maintenance monitoring method based on artificial intelligence, the step of optimizing the pre-built initial data mining neural network based on the multiple typical grid operation data and the key information mining results of the extended grid operation data corresponding to the at least one typical grid operation data to form the optimized data mining neural network corresponding to the initial data mining neural network includes:

[0038] Based on the typical grid operation data and the key information mining results of the extended grid operation data, combine to form a first optimized combination of grid data and / or a second optimized combination of grid data. The grid operation data sets corresponding to the typical grid operation data and the key information mining results of the extended grid operation data included in the first optimized combination of grid data are the same, and the grid operation data sets corresponding to the typical grid operation data and the key information mining results of the extended grid operation data included in the second optimized combination of grid data are different;

[0039] Based on the first optimized combination of grid data and / or the second optimized combination of grid data, optimize the pre-built initial data mining neural network to form the optimized data mining neural network corresponding to the initial data mining neural network.

[0040] In some preferred embodiments, in the above intelligent grid operation and maintenance monitoring method based on artificial intelligence, the step of analyzing and outputting the abnormal degree characterization value corresponding to the historical grid operation data according to the data correlation between the historical grid operation data and each reference grid operation data, and then determining the target operation safety degree of the target grid component based on the abnormal degree characterization value includes:

[0041] According to the data correlation between the historical grid operation data and each reference grid operation data, respectively determine the weight coefficient corresponding to each reference grid operation data. There may be a positive correlation between the weight coefficient and the data correlation;

[0042] Perform a weighted sum calculation on the abnormality degree reference values pre-configured for each of the reference power grid operation data according to the weight coefficient corresponding to each of the reference power grid operation data, so as to output the abnormality degree characterization value corresponding to the historical power grid operation data;

[0043] Determine the target operation safety degree of the target power grid component according to the abnormality degree characterization value, and there is a negative correlation between the target operation safety degree and the abnormality degree characterization value.

[0044] An embodiment of the present invention further provides an intelligent power grid operation and maintenance monitoring system based on artificial intelligence, including a processor and a memory. The memory is used to store a computer program, and the processor is used to execute the computer program to implement the above-mentioned intelligent power grid operation and maintenance monitoring method based on artificial intelligence.

[0045] An intelligent power grid operation and maintenance monitoring method and system based on artificial intelligence provided by an embodiment of the present invention can first perform data extraction processing on a target power grid component to obtain historical power grid operation data corresponding to the target power grid component; then calculate the data correlation degree between the historical power grid operation data and each reference power grid operation data included in a pre-configured reference power grid operation data set respectively; finally, based on the data correlation degree between the historical power grid operation data and each reference power grid operation data, and in combination with the abnormality degree reference value pre-configured for each reference power grid operation data, analyze and output the abnormality degree characterization value corresponding to the historical power grid operation data, and then determine the target operation safety degree of the target power grid component based on the abnormality degree characterization value. Based on this, since the operation safety analysis is performed on the historical power grid operation data of the entire target power grid component, rather than separately performing safety analysis on individual target power devices, and there is a correlation relationship between multiple target power devices, the basis for analysis is more sufficient, and the reliability of the operation safety analysis can be improved to a certain extent.

[0046] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following specifically gives preferred embodiments and cooperates with the attached drawings for detailed description as follows. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a structural block diagram of an intelligent power grid operation and maintenance monitoring system based on artificial intelligence provided by an embodiment of the present invention.

[0048] Figure 2 It is a schematic flow chart of each step included in an intelligent power grid operation and maintenance monitoring method based on artificial intelligence provided by an embodiment of the present invention.

[0049] Figure 3 It is a schematic diagram of each module included in an intelligent power grid operation and maintenance monitoring device based on artificial intelligence provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0050] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are only some, rather than all, of the embodiments of the present invention. Components of the embodiments of the present invention described and illustrated herein generally may be arranged and designed in a variety of different configurations. Therefore, the detailed description of the embodiments of the present invention provided herein is not intended to limit the scope of the claimed invention, but is merely representative of selected embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0051] As Figure 1 shown, the embodiments of the present invention provide an intelligent power grid operation and maintenance monitoring system based on artificial intelligence. Among them, the intelligent power grid operation and maintenance monitoring system may include a memory and a processor.

[0052] It should be understood that in some realizable embodiments, the memory and the processor are electrically connected directly or indirectly to achieve data transmission or interaction. For example, they may be electrically connected through one or more communication buses or signal lines. The memory may store at least one software functional module (computer program) that can exist in the form of software or firmware. The processor may be configured to execute the executable computer program stored in the memory, thereby implementing the intelligent power grid operation and maintenance monitoring method provided by the embodiments of the present invention.

[0053] It should be understood that in some implementable embodiments, the memory may be, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), etc. The processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), a system on chip (SoC), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0054] It should be understood that in some implementable embodiments, Figure 1 The structure shown is only illustrative, and the artificial intelligence-based intelligent power grid operation and maintenance monitoring system may further include more or fewer components than those shown in Figure 1 or have a configuration different from that shown in Figure 1 For example, it may include a communication unit for information interaction with other devices. It should be understood that in some implementable embodiments, the artificial intelligence-based intelligent power grid operation and maintenance monitoring system may be a server with data processing capabilities, or it may also be a server cluster formed by combining servers, etc.

[0055] Combined with Figure 2 , the embodiments of the present invention further provide an artificial intelligence-based intelligent power grid operation and maintenance monitoring method, which can be applied to the above-mentioned artificial intelligence-based intelligent power grid operation and maintenance monitoring system. Among them, the method steps defined by the processes related to the artificial intelligence-based intelligent power grid operation and maintenance monitoring method can be implemented by the artificial intelligence-based intelligent power grid operation and maintenance monitoring system.

[0056] Next, the specific process shown in Figure 2 will be elaborated in detail.

[0057] Step S110, perform data extraction processing on the target power grid component to obtain the historical power grid operation data corresponding to the target power grid component.

[0058] In an embodiment of the present invention, the artificial intelligence-based intelligent power grid operation and maintenance monitoring system can perform data extraction processing on a target power grid component to obtain historical power grid operation data corresponding to the target power grid component. The historical power grid operation data includes historical power parameters of each target power device in the target power grid component at multiple historical time points. The target power grid component includes a plurality of target power devices, and there is a correlation between the plurality of target power devices (this correlation may refer to a relationship in power operation, such as an electrical connection relationship, or a correlation between operation parameters, such as a following relationship in current change. In addition, the specific type of the historical power parameter is not limited. For example, it may be a current value, etc.).

[0059] Step S120: Calculate the data correlation degree between the historical power grid operation data and each reference power grid operation data included in a pre-configured reference power grid operation data set respectively.

[0060] In an embodiment of the present invention, the artificial intelligence-based intelligent power grid operation and maintenance monitoring system can calculate the data correlation degree between the historical power grid operation data and each reference power grid operation data included in a pre-configured reference power grid operation data set respectively (the reference power grid operation data may be power grid operation data in cases such as severe faults, general faults, minor faults, normal conditions, etc., and the abnormal degree reference values corresponding to the power grid operation data in various cases are different).

[0061] Step S130: Based on the data correlation degree between the historical power grid operation data and each reference power grid operation data, and in combination with the abnormal degree reference value pre-configured for each reference power grid operation data, analyze and output an abnormal degree characterization value corresponding to the historical power grid operation data, and then determine the target operation safety degree corresponding to the target power grid component based on the abnormal degree characterization value.

[0062] In an embodiment of the present invention, the artificial intelligence-based intelligent power grid operation and maintenance monitoring system can, based on the data correlation degree between the historical power grid operation data and each reference power grid operation data, and in combination with the abnormal degree reference value pre-configured for each reference power grid operation data, analyze and output an abnormal degree characterization value corresponding to the historical power grid operation data, and then determine the target operation safety degree corresponding to the target power grid component based on the abnormal degree characterization value.

[0063] Based on this, since the operation safety analysis is performed on the historical power grid operation data of the entire target power grid component rather than separately on each individual target power device, and there is a correlation between the multiple target power devices, the basis for the analysis is more sufficient, which can improve the reliability of the operation safety analysis to a certain extent, thereby improving the deficiencies in the prior art.

[0064] It should be understood that in the process of executing the above step S110, in some implementable embodiments, the following sub-steps may be specifically executed:

[0065] According to the historical time period (such as a duration of 10 days, 20 days, one month, etc. going back from the current moment), data extraction processing is performed on the target power grid component to obtain the initial historical power grid operation data corresponding to the target power grid component, and the initial historical power grid operation data includes data of each time point of the target power grid component within the historical time period;

[0066] Sample the historical time period to form multiple historical time points, and the length of the historical time interval between the multiple historical time points has a decreasing corresponding relationship along the time trend (for example, data 10 days ago can be sampled once every hour, data 5 days ago can be sampled once every 0.5 hours, and data of the most recent day can be sampled once every 5 minutes);

[0067] For each of the historical time points, from the initial historical power grid operation data, the historical power parameters of each target power device in the target power grid component are respectively extracted at this historical time point to obtain the historical power grid operation data corresponding to the target power grid component.

[0068] It should be understood that in the process of executing the above step S120, in some implementable embodiments, the following sub-steps may be specifically executed:

[0069] Extract a reference power grid operation data from the multiple reference power grid operation data included in the pre-configured reference power grid operation data set, and then use this reference power grid operation data and the historical power grid operation data as a data combination for calculating the data relevance (based on this, corresponding multiple data combinations can be formed for multiple reference power grid operation data);

[0070] Load the historical power grid operation data and the reference power grid operation data included in the data combination into the optimized data mining neural network respectively, and use the optimized data mining neural network to respectively mine the key information mining results of the power grid operation data corresponding to the historical power grid operation data and the key information mining results of the power grid operation data corresponding to the reference power grid operation data (the key information mining function of the optimized data mining neural network can be obtained through network optimization, and its optimization process can refer to the relevant description later);

[0071] Based on the key information mining results of the historical power grid operation data and the key information mining results of the reference power grid operation data, calculate and output the matching degree of the key information mining results corresponding to the data combination, as the data correlation between the historical power grid operation data and the reference power grid operation data included in the data combination (exemplarily, when the key information mining results of the power grid operation data and the key information mining results of the power grid operation data are in the form of vectors, the vector distance between the key information mining results of the power grid operation data and the key information mining results of the power grid operation data can be calculated, and then the corresponding data correlation can be calculated based on the analyzed vector distance, such as negative correlation between the two; in addition, the optimized data mining neural network may include a feature mining model to perform feature mining, so as to obtain the key information mining results represented by vectors).

[0072] It should be understood that in the process of executing the above step S120, in order to facilitate the optimized data mining neural network to effectively mine key information, the historical power grid operation data and the reference power grid operation data can also be preprocessed, so as to load the preprocessed historical power grid operation data and the preprocessed reference power grid operation data into the optimized data mining neural network respectively, and use the optimized data mining neural network to respectively mine the corresponding key information mining results of the power grid operation data and the corresponding key information mining results of the power grid operation data. In some implementable embodiments, the preprocessing may specifically include the following sub-steps:

[0073] Classify the historical power parameters included in the historical power grid operation data according to the corresponding target power devices, so as to obtain multiple historical power parameter sets corresponding to the multiple target power devices, and then serialize the historical power parameters included in each historical power parameter set according to the corresponding historical time points, so as to form a historical power parameter sequence corresponding to each target power device, and multiple historical power parameter sequences corresponding to multiple target power devices;

[0074] For each historical power parameter sequence, based on the historical power parameter sequence, construct a corresponding historical power parameter change amplitude sequence. In the historical power parameter change amplitude sequence, each historical power parameter change amplitude is equal to the difference between two adjacent historical power parameters at the corresponding sequence positions in the historical power parameter sequence (such as the latter minus the former, etc.);

[0075] Based on the corresponding historical time points, two-dimensional spatial projection processing is respectively performed on each historical power parameter change amplitude included in each of the historical power parameter change amplitude sequences to form two-dimensional spatial coordinate points corresponding to each of the historical power parameter change amplitudes. Then, every two adjacent two-dimensional spatial coordinate points are connected by a straight line segment to form a straight line segment connection path corresponding to each of the historical power parameter change amplitude sequences. Moreover, curve fitting processing is performed on the two-dimensional spatial coordinate points to form a curved path corresponding to the historical power parameter change amplitude sequence;

[0076] For each of the curved paths, the difference between the historical power parameter change amplitudes corresponding to every two adjacent peak points in the curved path is calculated respectively to obtain the difference in historical power parameter change amplitudes between every two adjacent peak points. Then, based on the comparison between the difference in historical power parameter change amplitudes and a pre-configured difference threshold, when the difference in historical power parameter change amplitudes is greater than or equal to the difference threshold, the trough point between the two adjacent peak points corresponding to the difference in historical power parameter change amplitudes is used as a segmentation point. Moreover, based on the segmentation point, the curved path is segmented to form a plurality of segmented curved path segments corresponding to the curved path. Furthermore, based on the plurality of segmented curved path segments, the straight line segment connection path corresponding to the curved path is segmented to form a corresponding number of segmented straight line segment connection path segments;

[0077] For each curve segment included in each of the segmented curved path segments, based on the corresponding historical time point, in the corresponding segmented straight line segment connection path segment, the straight line segment corresponding to the curve segment is determined, and then the curve segment is adjusted based on the straight line segment to form an adjusted curve segment corresponding to the curve segment. The tangent of the adjusted curve segment is parallel to the corresponding straight line segment (the specific adjustment method is not limited as long as the purpose of parallelism is satisfied);

[0078] Based on the corresponding adjusted curve segments, adjusted curved paths corresponding to each of the historical power parameter change amplitude sequences are respectively formed. Based on the adjusted curved paths corresponding to the historical power parameter change amplitude sequences, the sequence correlation degree between every two of the multiple historical power parameter sequences is calculated to output the sequence correlation degree between every two historical power parameter sequences (exemplarily, the path coincidence degree between the corresponding two adjusted curved paths can be used as the sequence correlation degree between the corresponding two historical power parameter sequences);

[0079] Sort the multiple historical power parameter sequences according to the corresponding sequence correlation degree to form an ordered set of historical power parameter sequences corresponding to the multiple historical power parameter sequences (exemplarily, in the ordered set of historical power parameter sequences, the average value of the sequence correlation degrees between every two adjacent historical power parameter sequences can have the maximum value in various sorts), and network process the ordered set of historical power parameter sequences to form a corresponding historical power parameter distribution network, and this historical power parameter distribution network is used as the preprocessed historical power grid operation data (exemplarily, the process of preprocessing the reference power grid operation data can be the same as the above process; in addition, in the historical power parameter distribution network, the network distribution coordinates of each historical power parameter are determined based on the set position of the corresponding historical power parameter sequence in the ordered set of historical power parameter sequences and the corresponding historical time point).

[0080] It should be understood that in some implementable embodiments, the optimization process of the optimized data mining neural network can specifically perform the following sub-steps:

[0081] Extract multiple typical power grid operation data, and extract the grid operation data identification information corresponding to the multiple typical power grid operation data (the multiple typical power grid operation data and the grid operation data identification information corresponding to the multiple typical power grid operation data are used as the basis for optimizing the initial data mining neural network to enable it to have the function of mining key data information);

[0082] Based on the mining results of the key information of the grid operation data corresponding to at least one typical power grid operation data and the corresponding grid operation data identification information, mine the mining results of the key information of the extended grid operation data corresponding to the at least one typical power grid operation data;

[0083] Optimize the pre-built initial data mining neural network according to the multiple typical power grid operation data and the mining results of the key information of the extended grid operation data corresponding to the at least one typical power grid operation data to form a corresponding optimized data mining neural network (in this way, on the basis of using the multiple typical power grid operation data as the basis for optimization, by mining the mining results of the key information of the extended grid operation data corresponding to the at least one typical power grid operation data, which is also used as the basis for optimization, it is possible to reduce the optimization cost to a certain extent and improve the richness of the optimized data, so that the mining function of the optimized data mining neural network is more reliable).

[0084] It should be understood that, in the process of performing the steps of extracting a plurality of typical power grid operation data and extracting the power grid operation data identification information corresponding to the plurality of typical power grid operation data, in some implementable embodiments, the following sub-steps may be specifically performed:

[0085] Extract a plurality of typical power grid operation data, and screen out at least one first typical power grid operation data from the plurality of typical power grid operation data, where the at least one first typical power grid operation data belongs to the typical power grid operation data in the plurality of typical power grid operation data that does not have actual power grid operation data identification information; and, based on the mining results of the key information of the power grid operation data corresponding to the at least one first typical power grid operation data, classify and form corresponding at least one classification identification information (for example, the first typical power grid operation data with a relatively high matching degree between the corresponding mining results of the key information of the power grid operation data can be assigned to a classification set, and then a corresponding classification identification information, such as a number, etc., can be configured for this classification set); and, for each of the first typical power grid operation data, determine the power grid operation data identification information corresponding to the first typical power grid operation data according to the classification identification information corresponding to the first typical power grid operation data (exemplarily, the power grid operation data identification information is used to distinguish whether the corresponding typical power grid operation data is similar; in addition, the above-mentioned having actual power grid operation data identification information may be manually marked).

[0086] It should be understood that, in the process of performing the step of mining the mining results of the extended key information of the power grid operation data corresponding to the at least one typical power grid operation data based on the mining results of the key information of the power grid operation data corresponding to the at least one typical power grid operation data and the corresponding power grid operation data identification information, in some implementable embodiments, the following sub-steps may be specifically performed:

[0087] Based on the key information mining results of power grid operation data corresponding to at least one typical power grid operation data and the corresponding power grid operation data identification information, the first key information mining results of power grid operation data corresponding to the at least one typical power grid operation data are mined; and, based on the first key information mining results of power grid operation data corresponding to the at least one typical power grid operation data, the second key information mining results of power grid operation data corresponding to the at least one typical power grid operation data are analyzed and output; and, based on the key information mining results of power grid operation data corresponding to the at least one typical power grid operation data, the corresponding first key information mining results of power grid operation data and the corresponding second key information mining results of power grid operation data, the extended key information mining results of power grid operation data corresponding to the at least one typical power grid operation data are determined (based on this, the extended key information mining results of power grid operation data corresponding to each typical power grid operation data in the at least one typical power grid operation data can be generated through the corresponding three key information mining results).

[0088] It should be understood that in the process of performing the step of mining the first key information mining results of power grid operation data corresponding to the at least one typical power grid operation data based on the key information mining results of power grid operation data corresponding to the at least one typical power grid operation data and the corresponding power grid operation data identification information, in some implementable embodiments, the following sub-steps may be specifically executed:

[0089] Based on the power grid operation data identification information corresponding to at least one typical power grid operation data, the typical power grid operation data with the same corresponding power grid operation data identification information are classified into a power grid operation data set to form at least one power grid operation data set;

[0090] Based on the key information mining results of power grid operation data corresponding to the typical power grid operation data included in each of the power grid operation data sets, the key information mining results of the power grid operation data sets corresponding to each of the power grid operation data sets are (respectively) mined (exemplarily, for each power grid operation data set, the mean value calculation can be performed on the key information mining results of power grid operation data corresponding to the typical power grid operation data included in the power grid operation data set to obtain the key information mining results of the power grid operation data set corresponding to the power grid operation data set; in addition, either the existing feature mining model can be used to mine the key information mining results of power grid operation data corresponding to the at least one typical power grid operation data, or the initial data mining neural network can be used to mine the key information mining results of power grid operation data corresponding to the at least one typical power grid operation data; in addition, when there are different optimization batches, for the optimization process of the current batch, the initial data mining neural network optimized in the previous batch can be used for information mining);

[0091] Mark the key information mining result of the power grid operation data set corresponding to the at least one typical power grid operation data as the first key information mining result of the power grid operation data pair corresponding to the at least one typical power grid operation data.

[0092] It should be understood that in the process of performing the step of analyzing and outputting the second key information mining result of the power grid operation data corresponding to the at least one typical power grid operation data based on the first key information mining result of the power grid operation data corresponding to the at least one typical power grid operation data, in some implementable embodiments, the following sub-steps may be specifically executed:

[0093] Filter out the power grid operation data set to be confirmed corresponding to the at least one typical power grid operation data (exemplarily, at least two other power grid operation data sets other than the power grid operation data set corresponding to the typical power grid operation data may be used as the power grid operation data set to be confirmed corresponding to the typical power grid operation data, which may be any two, or may be filtered based on the correlation between the power grid operation data identification information corresponding to the typical power grid operation data included in the other power grid operation data sets and the power grid operation data identification information corresponding to the typical power grid operation data);

[0094] Based on the first key information mining result of the power grid operation data corresponding to the at least one typical power grid operation data and the key information mining result of the power grid operation data set corresponding to the power grid operation data set to be confirmed, analyze the confirmed power grid operation data set corresponding to the at least one typical power grid operation data in the power grid operation data set to be confirmed (exemplarily, the matching degree between the first key information mining result of the power grid operation data and the key information mining result of the power grid operation data set corresponding to the power grid operation data set to be confirmed may be calculated first, and then, the power grid operation data set to be confirmed with a matching degree greater than or equal to the matching degree reference value may be marked to form the corresponding confirmed power grid operation data set. In addition, the confirmed power grid operation data set may be one or more, and no specific limitation is made thereto);

[0095] Mark the key information mining result of the power grid operation data set corresponding to the confirmed power grid operation data set corresponding to the at least one typical power grid operation data as the second key information mining result of the power grid operation data corresponding to the at least one typical power grid operation data.

[0096] It should be understood that in the process of performing the step of determining the key information mining result of the extended grid operation data corresponding to the at least one typical grid operation data based on the key information mining result of the grid operation data corresponding to the at least one typical grid operation data, the corresponding first key information mining result of the grid operation data, and the corresponding second key information mining result of the grid operation data, in some implementable embodiments, the following sub-steps may be specifically performed:

[0097] Determine a target variable, and there is a positive correlation between the current value of the target variable and the optimization progress of optimizing the initial data mining neural network (that is, the current value of the target variable corresponding to the first optimization can be less than the current value of the target variable corresponding to the second optimization, the current value of the target variable corresponding to the second optimization can be less than the current value of the target variable corresponding to the third optimization, the current value of the target variable corresponding to the third optimization can be less than the current value of the target variable corresponding to the fourth optimization, and so on);

[0098] Based on the key information mining result of the grid operation data corresponding to the at least one typical grid operation data, the corresponding first key information mining result of the grid operation data, the corresponding second key information mining result of the grid operation data, and the target variable, determine the key information mining result of the extended grid operation data corresponding to the at least one typical grid operation data (since the first key information mining result of the grid operation data can represent the key information of the grid operation data set corresponding to the typical grid operation data, the second key information mining result of the grid operation data can represent the key information of other grid operation data sets similar to the grid operation data set corresponding to the typical grid operation data, and by integrating the key information mining result of the grid operation data of the typical grid operation data itself, a key information mining result with richer information can be formed, that is, the key information mining result of the extended grid operation data of the typical grid operation data; exemplarily, the sum value between the second key information mining result of the grid operation data and the first key information mining result of the grid operation data can be calculated first, then, the sum value is weighted based on the target variable, and then the weighted result is superimposed with the key information mining result of the grid operation data; in other embodiments, the key information mining result of the grid operation data corresponding to the at least one typical grid operation data, the corresponding first key information mining result of the grid operation data, and the corresponding second key information mining result of the grid operation data can also be directly superimposed).

[0099] It should be understood that in the process of performing the step of optimizing the pre-established initial data mining neural network based on the key information mining results of the multiple typical power grid operation data and the extended power grid operation data corresponding to the at least one typical power grid operation data to form the optimized data mining neural network corresponding to the initial data mining neural network, in some implementable embodiments, the following sub-steps may be specifically executed:

[0100] Based on the typical power grid operation data and the key information mining results of the extended power grid operation data, a first power grid data optimization combination and / or a second power grid data optimization combination are formed. The power grid operation data sets corresponding to the typical power grid operation data and the key information mining results of the extended power grid operation data included in the first power grid data optimization combination are consistent, and the power grid operation data sets corresponding to the typical power grid operation data and the key information mining results of the extended power grid operation data included in the second power grid data optimization combination are inconsistent;

[0101] Based on the first power grid data optimization combination and / or the second power grid data optimization combination, the pre-established initial data mining neural network is optimized to form the optimized data mining neural network corresponding to the initial data mining neural network (that is to say, the pre-established initial data mining neural network can be optimized only based on the first power grid data optimization combination, or only based on the second power grid data optimization combination, or also based on both the first power grid data optimization combination and the second power grid data optimization combination; wherein, the optimization process may be to optimize the network parameters of the initial data mining neural network so that the error obtained based on the optimized parameters converges, or the number of optimization times reaches a reference number, etc. This error may refer to the degree of difference between the key information mining results of the extended power grid operation data and the key information mining results of the data mined by the initial data mining neural network for the typical power grid operation data).

[0102] It should be understood that in the process of performing step S130 above, in some implementable embodiments, the following sub-steps may be specifically executed:

[0103] According to the data correlation between the historical power grid operation data and each reference power grid operation data, the weight coefficient corresponding to each reference power grid operation data is determined respectively. There may be a positive correlation corresponding relationship between the weight coefficient and the data correlation (exemplarily, the sum value of the weight coefficients corresponding to each reference power grid operation data may be equal to 1);

[0104] Based on the weight coefficient corresponding to each of the reference power grid operation data, perform a weighted summation calculation on the reference abnormal degree values pre-configured for each of the reference power grid operation data, so as to output the abnormal degree characterization value (i.e., the weighted summation value) corresponding to the historical power grid operation data;

[0105] Determine the target operation safety degree of the target power grid component according to the abnormal degree characterization value, and there is a negative correlation between the target operation safety degree and the abnormal degree characterization value (that is, the larger the abnormal degree characterization value, the lower the target operation safety degree).

[0106] Combined with Figure 3 , an embodiment of the present invention further provides an intelligent power grid operation and maintenance monitoring device based on artificial intelligence, which can be applied to the above-mentioned intelligent power grid operation and maintenance monitoring system based on artificial intelligence. Among them, the intelligent power grid operation and maintenance monitoring device based on artificial intelligence may include software function modules such as a historical power grid operation data extraction module, a data correlation calculation module, and an operation safety analysis module.

[0107] The historical power grid operation data extraction module is used to perform data extraction processing on the target power grid component to obtain the historical power grid operation data corresponding to the target power grid component. The historical power grid operation data includes the historical power parameters of each target power device in the target power grid component at multiple historical time points. The target power grid component includes multiple target power devices, and there is a correlation between the multiple target power devices; the data correlation calculation module is used to calculate the data correlation between the historical power grid operation data and each reference power grid operation data included in the pre-configured reference power grid operation data set; the operation safety analysis module is used to analyze and output the abnormal degree characterization value corresponding to the historical power grid operation data based on the data correlation between the historical power grid operation data and each reference power grid operation data, and in combination with the abnormal degree reference value pre-configured for each reference power grid operation data, and then determine the target operation safety degree of the target power grid component based on the abnormal degree characterization value.

[0108] In summary, for the intelligent power grid operation and maintenance monitoring method and system based on artificial intelligence provided by the present invention, data extraction and processing can be first performed on target power grid components to obtain historical power grid operation data corresponding to the target power grid components; then, the data correlation degrees between the historical power grid operation data and each reference power grid operation data included in a pre-configured set of reference power grid operation data are calculated respectively; finally, based on the data correlation degrees between the historical power grid operation data and each reference power grid operation data, and in combination with the abnormal degree reference values pre-configured for each reference power grid operation data, an abnormal degree characterization value corresponding to the historical power grid operation data is analyzed and output, and then the target operation safety degree corresponding to the target power grid components is determined based on the abnormal degree characterization value. Based on this, since the operation safety analysis is performed on the historical power grid operation data of the overall target power grid components rather than on individual target power devices separately, and there is a correlation relationship between multiple target power devices, the basis for the analysis is more sufficient, and the reliability of the operation safety analysis can be improved to a certain extent.

[0109] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. An intelligent grid operation and maintenance monitoring method based on artificial intelligence, characterized in that, Including: Performing data extraction processing on a target power grid component to obtain historical power grid operation data corresponding to the target power grid component, where the historical power grid operation data includes historical power parameters of each target power device in the target power grid component at multiple historical time points, the target power grid component includes multiple target power devices, and there is a correlation relationship between the multiple target power devices; Calculating the data correlation degree between the historical power grid operation data and each reference power grid operation data included in a pre-configured reference power grid operation data set respectively; Based on the data correlation degree between the historical power grid operation data and each reference power grid operation data, and in combination with the abnormal degree reference value pre-configured for each reference power grid operation data, analyzing and outputting an abnormal degree characterization value corresponding to the historical power grid operation data, and then determining the target operation safety degree corresponding to the target power grid component based on the abnormal degree characterization value; The step of calculating the data correlation degree between the historical power grid operation data and each reference power grid operation data included in the pre-configured reference power grid operation data set respectively includes: Extracting a reference power grid operation data from the multiple reference power grid operation data included in the pre-configured reference power grid operation data set, and then using the reference power grid operation data and the historical power grid operation data as a data combination for which the data correlation degree is to be calculated; Loading the historical power grid operation data and the reference power grid operation data included in the data combination into an optimized data mining neural network respectively, and using the optimized data mining neural network to respectively mine out the key information mining result of the power grid operation data corresponding to the historical power grid operation data and the key information mining result of the power grid operation data corresponding to the reference power grid operation data; Based on the key information mining result of the power grid operation data corresponding to the historical power grid operation data and the key information mining result of the power grid operation data corresponding to the reference power grid operation data, calculating and outputting the matching degree of the key information mining result corresponding to the data combination as the data correlation degree between the historical power grid operation data and the reference power grid operation data included in the data combination.

2. The intelligent power grid operation and maintenance monitoring method based on artificial intelligence according to claim 1, characterized in that, The step of performing data extraction processing on the target power grid component to obtain the historical power grid operation data corresponding to the target power grid component includes: Performing data extraction processing on the target power grid component according to a historical time period to obtain initial historical power grid operation data corresponding to the target power grid component, where the initial historical power grid operation data includes data at each time point within the historical time period of the target power grid component; Sampling the historical time period to form multiple historical time points, and there is a decreasing corresponding relationship between the lengths of the historical time intervals between the multiple historical time points along the time trend; For each historical time point, respectively extracting the historical power parameters of each target power device in the target power grid component from the initial historical power grid operation data to obtain the historical power grid operation data corresponding to the target power grid component.

3. The intelligent power grid operation and maintenance monitoring method based on artificial intelligence according to claim 1, wherein, The optimization process of the optimized data mining neural network includes: Extracting a plurality of typical power grid operation data, and extracting the power grid operation data identification information corresponding to the plurality of typical power grid operation data; Based on the power grid operation data key information mining results corresponding to at least one typical power grid operation data and the corresponding power grid operation data identification information, mining the extended power grid operation data key information mining results corresponding to the at least one typical power grid operation data; Optimizing the pre-built initial data mining neural network according to the plurality of typical power grid operation data and the extended power grid operation data key information mining results corresponding to the at least one typical power grid operation data to form the corresponding optimized data mining neural network; Among them, the step of extracting a plurality of typical power grid operation data and extracting the power grid operation data identification information corresponding to the plurality of typical power grid operation data includes: Extracting a plurality of typical power grid operation data, and screening out at least one first typical power grid operation data from the plurality of typical power grid operation data, where the at least one first typical power grid operation data belongs to the typical power grid operation data without actual power grid operation data identification information among the plurality of typical power grid operation data; and, classifying and forming the corresponding at least one classification identification information based on the power grid operation data key information mining results corresponding to the at least one first typical power grid operation data; and, for each of the first typical power grid operation data, determining the power grid operation data identification information corresponding to the first typical power grid operation data according to the classification identification information corresponding to the first typical power grid operation data.

4. The intelligent power grid operation and maintenance monitoring method based on artificial intelligence according to claim 3, characterized in that The step of mining the extended power grid operation data key information mining results corresponding to the at least one typical power grid operation data based on the power grid operation data key information mining results corresponding to the at least one typical power grid operation data and the corresponding power grid operation data identification information includes: Mining the first power grid operation data key information mining results corresponding to the at least one typical power grid operation data based on the power grid operation data key information mining results corresponding to the at least one typical power grid operation data and the corresponding power grid operation data identification information; Analyzing and outputting the second power grid operation data key information mining results corresponding to the at least one typical power grid operation data based on the first power grid operation data key information mining results corresponding to the at least one typical power grid operation data; Determining the extended power grid operation data key information mining results corresponding to the at least one typical power grid operation data based on the power grid operation data key information mining results corresponding to the at least one typical power grid operation data, the corresponding first power grid operation data key information mining results, and the corresponding second power grid operation data key information mining results; Among them, the step of mining the first power grid operation data key information mining results corresponding to the at least one typical power grid operation data based on the power grid operation data key information mining results corresponding to the at least one typical power grid operation data and the corresponding power grid operation data identification information includes: Based on the grid operation data identification information corresponding to at least one typical grid operation data, the typical grid operation data with the same corresponding grid operation data identification information is attributed to a grid operation data set to form at least one grid operation data set; based on the grid operation data key information mining results corresponding to the typical grid operation data included in each of the grid operation data sets, the grid operation data set key information mining results corresponding to each of the grid operation data sets are mined; the grid operation data set key information mining results corresponding to the grid operation data sets corresponding to the at least one typical grid operation data are marked as the first grid operation data key information mining results of the at least one pair of typical grid operation data.

5. The intelligent grid operation and maintenance monitoring method based on artificial intelligence according to claim 4, characterized in that, The step of analyzing and outputting the second grid operation data key information mining results corresponding to the at least one typical grid operation data based on the first grid operation data key information mining results corresponding to the at least one typical grid operation data includes: Filtering out the to-be-confirmed grid operation data sets corresponding to the at least one typical grid operation data; and, based on the first grid operation data key information mining results corresponding to the at least one typical grid operation data and the grid operation data set key information mining results corresponding to the to-be-confirmed grid operation data sets, analyzing out the confirmed grid operation data sets corresponding to the at least one typical grid operation data in the to-be-confirmed grid operation data sets; Marking the grid operation data set key information mining results corresponding to the confirmed grid operation data sets corresponding to the at least one typical grid operation data as the second grid operation data key information mining results corresponding to the at least one typical grid operation data.

6. The intelligent power grid operation and maintenance monitoring method based on artificial intelligence according to claim 4, characterized in that, The step of determining the extended grid operation data key information mining results corresponding to the at least one typical grid operation data based on the grid operation data key information mining results, the corresponding first grid operation data key information mining results, and the corresponding second grid operation data key information mining results corresponding to the at least one typical grid operation data includes: Determining a target variable, where there is a positive correlation between the current value of the target variable and the optimization progress of optimizing the initial data mining neural network; Based on the grid operation data key information mining results, the corresponding first grid operation data key information mining results, the corresponding second grid operation data key information mining results, and the target variable corresponding to the at least one typical grid operation data, determining the extended grid operation data key information mining results corresponding to the at least one typical grid operation data.

7. The intelligent power grid operation and maintenance monitoring method based on artificial intelligence according to claim 3, characterized in that The step of optimizing the pre-built initial data mining neural network according to the multiple typical grid operation data and the extended grid operation data key information mining results corresponding to the at least one typical grid operation data to form the optimized data mining neural network corresponding to the initial data mining neural network includes: Based on the typical power grid operation data and the key information mining results of the extended power grid operation data, a first optimized combination of power grid data and / or a second optimized combination of power grid data is formed. The power grid operation data sets corresponding to the typical power grid operation data and the key information mining results of the extended power grid operation data included in the first optimized combination of power grid data are consistent, and the power grid operation data sets corresponding to the typical power grid operation data and the key information mining results of the extended power grid operation data included in the second optimized combination of power grid data are inconsistent; Based on the first optimized combination of power grid data and / or the second optimized combination of power grid data, the initially established initial data mining neural network is optimized to form an optimized data mining neural network corresponding to the initial data mining neural network.

8. The intelligent power grid operation and maintenance monitoring method based on artificial intelligence according to any one of claims 1-7, characterized in that, The step of analyzing and outputting the abnormal degree characterization value corresponding to the historical power grid operation data according to the data correlation between the historical power grid operation data and each of the reference power grid operation data, and combining the abnormal degree reference value pre-configured for each of the reference power grid operation data, and then determining the target operation safety degree of the target power grid component based on the abnormal degree characterization value includes: According to the data correlation between the historical power grid operation data and each of the reference power grid operation data, the weight coefficient corresponding to each of the reference power grid operation data is determined respectively, and there is a positive correlation correspondence between the weight coefficient and the data correlation; According to the weight coefficient corresponding to each of the reference power grid operation data, the weighted sum calculation is performed on the abnormal degree reference value pre-configured for each of the reference power grid operation data to output the abnormal degree characterization value corresponding to the historical power grid operation data; The target operation safety degree of the target power grid component is determined according to the abnormal degree characterization value, and there is a negative correlation between the target operation safety degree and the abnormal degree characterization value.

9. An intelligent power grid operation and maintenance monitoring system based on artificial intelligence, characterized in that, It includes a processor and a memory. The memory is used to store a computer program, and the processor is used to execute the computer program to implement the method according to any one of claims 1-8.

Images