Operation inspection analysis method based on mechanism and data fusion driving, medium and system
By building multiple models of power equipment and carrying out weighted fusion, combining time series analysis and machine learning, accurate evaluation of power equipment status and accurate prediction of future situations are achieved, solving the problems of inaccurate evaluation and low prediction accuracy in the prior art.
Patent Information
- Application Number
- CN202411819121.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-11
- Publication Date
- 2025-04-29
AI Technical Summary
The existing power grid operation and inspection status analysis methods cannot conduct effective quantitative evaluation, and the accuracy of power equipment status prediction is not ideal and the accuracy is low.
Using the operation and inspection analysis method driven by mechanism and data fusion, a digital geometric model, physical model, behavioral model and rule model of power equipment is constructed. The outputs of each model are fused through a weighted fusion algorithm to obtain comprehensive state evaluation results, and time series analysis and machine learning methods are used to predict the future state change trend of the equipment.
It realizes accurate assessment of the status of power equipment and accurate prediction of future situations, and improves the accuracy of power operation and inspection risk identification at component level and equipment level under complex operating conditions.
Smart Images
Figure CN119903448A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of power data analysis, and in particular to an operation and inspection analysis method, medium and system driven by mechanism and data fusion. Background Art
[0002] In recent years, with the rapid development of intelligent decision-making analysis, more and more fields have begun to use intelligent decision-making analysis to analyze equipment operation status data; among them, smart decision-making is a means of comprehensive analysis of equipment operation status data based on relevant standards, algorithms and experience. It can determine the status and change trend of the equipment, and predict the remaining life and operating status of the equipment, thereby providing a basis for the formulation of subsequent operation and maintenance strategies for the equipment.
[0003] In the current power field, the commonly used state prediction and assessment methods for power transmission and transformation equipment are: 1. Guideline-based condition assessment method; The guideline-based condition assessment method is the main basis for equipment condition assessment at present. The assessor deducts points from the equipment condition quantity according to the inspection and test results and the deduction standards for the equipment condition quantity specified in the guideline to determine the equipment condition level. This method is simple and easy to use and widely used. However, this guideline-based condition assessment method only deducts points according to a simple equipment evaluation system, and cannot perform effective quantitative evaluation and condition prediction for complex defect situations.
[0004] 2. State assessment method based on classification and regression; Classification and regression methods mainly use a large amount of historical data as samples, and establish evaluation models by studying the mutual influence and change rules between different variables. They are usually used for prediction and pattern recognition. Common classification and regression methods include artificial neural networks, decision trees, support vector machines, etc.
[0005] However, the training of such models requires a large amount of equipment defect and fault status data; however, the reliability of power transmission and transformation equipment is high, and the defect and fault status data are scarce, resulting in insufficient training samples, resulting in less than ideal model accuracy; at the same time, some models converge slowly and are prone to fall into local optimality; in addition, this state assessment method cannot explain the physical connection between state parameters and equipment status from a mechanistic perspective, which is not conducive to the understanding and application of the assessment results. Summary of the invention
[0006] The technical problem to be solved by the present invention is: the present invention discloses an operation and inspection analysis method, medium and system driven by mechanism and data fusion, so as to solve the problems existing in the existing power grid operation and inspection status analysis method that effective quantitative evaluation cannot be carried out, the accuracy of power equipment status prediction is not ideal and the precision is low.
[0007] In order to solve the above technical problems, the technical solution adopted by the present invention is: an operation and inspection analysis method based on mechanism and data fusion drive, which comprises the steps of: S1: Construct digital geometric model, physical model, behavioral model and rule model of power equipment; S2: collecting the real-time operation data of the power equipment and preprocessing it, so as to input the preprocessed real-time operation data into each model and obtain the output of each model; S3: using a weighted fusion algorithm to fuse the outputs of each model to obtain a comprehensive status assessment result of the power equipment; S4: Based on the comprehensive status assessment results, use time series analysis and machine learning methods to predict future status change trends of the power equipment.
[0008] In the present invention, a new operation and inspection analysis method based on mechanism and data fusion is designed. The operation and inspection analysis method constructs a digital geometric model, a physical model, a behavioral model and a rule model of the power equipment to respectively describe the physical structure, multi-physical field behavior, dynamic behavior and empirical rules of the equipment; and pre-processes the real-time operation data of the power equipment and inputs it into each model, and then combines the weighted fusion algorithm to fuse the outputs of each model to obtain a comprehensive state evaluation result, so as to predict the future state change trend of the power equipment based on the comprehensive state evaluation result.
[0009] In the present invention, the operation and maintenance analysis method based on mechanism and data fusion drive studies the multi-physical field mechanism numerical simulation technology and artificial intelligence algorithm analysis technology, and establishes a high-approximation state assessment and situation prediction method driven by "model + data" fusion in the power digital space. It can effectively improve the accuracy of component-level and equipment-level power operation and maintenance risk identification under complex working conditions, that is, it can effectively solve the problems of existing power grid analysis methods that cannot perform effective quantitative evaluation, and the accuracy and low precision of power equipment state prediction are not ideal.
[0010] The operation and maintenance analysis method of the present invention develops synchronous evolution and adaptive fusion mechanism analysis technology of mechanism simulation model and data-driven, constructs high-approximation simulation analysis driven by "model + data" fusion of digital geometric model, physical model, behavior model and rule model, establishes a two-layer iterative intelligent predictive operation and maintenance model driven by "model + data" fusion, realizes diagnosis, prediction and auxiliary decision-making of power equipment under complex working conditions, and has good promotion prospects and application value.
[0011] Further, in the operation and inspection analysis method based on mechanism and data fusion drive described in the present invention, in step S3, the comprehensive status evaluation result is specifically: S = αMg + βMp + γMb + δMr Among them, S represents the comprehensive state assessment result; Mg, Mp, Mb, and Mr represent the outputs of the digital geometry model, physical model, behavior model, and rule model, respectively; α, β, γ, and δ represent different weight coefficients, respectively.
[0012] Further, in the operation and inspection analysis method based on mechanism and data fusion drive described in the present invention, in step S4, a situation prediction model is obtained by using time series analysis and machine learning methods to predict the future state change trend of the power equipment based on the situation prediction model, and the situation prediction model is specifically: P(t)=f(S,t) Wherein, P(t) is the situation prediction model, f is the prediction function, S represents the comprehensive state evaluation result, t is the time variable, and f(S, t) represents the equipment state prediction of the power equipment at a specific time point t.
[0013] Furthermore, in the operation and inspection analysis method based on mechanism and data fusion drive described in the present invention, in step S1, an initial physical model and an initial behavioral model of the power equipment are constructed, and the initial physical model and the initial behavioral model are reduced in order through a singular value decomposition method and a principal component analysis method to obtain the physical model and the behavioral model.
[0014] Furthermore, in the operation and inspection analysis method based on mechanism and data fusion drive described in the present invention, in step S1, the initial physical model and the initial behavior model are reduced in order, and the specific steps are: Performing singular value decomposition on the initial physical model, retaining the first k singular values, and obtaining the physical model after reduction; Perform principal component analysis on the initial behavior model, retain the first i principal components, and obtain the reduced-order behavior model.
[0015] Furthermore, in the operation and inspection analysis method based on mechanism and data fusion drive described in the present invention, in step S1, an initial rule model of the power equipment is constructed, and the initial rule model is pruned through pruning technology to remove unimportant rule items to obtain a lightweight rule model.
[0016] Furthermore, in the operation and inspection analysis method based on mechanism and data fusion drive described in the present invention, in step S3, a weighted fusion algorithm is used to fuse the outputs of each model to obtain an initial comprehensive state assessment result of the power equipment, the initial comprehensive state assessment result is quantized, and the floating-point parameters of the initial comprehensive state assessment result are converted into integers to obtain the lightweight comprehensive state assessment result.
[0017] In the above technical scheme of the present invention, in order to improve the timeliness of prediction and calculation efficiency, the operation and inspection analysis method described in the present invention also applies model reduction and lightweight technology to solve the problem of long simulation and calculation time, thereby reducing the delay and improving the timeliness of prediction while ensuring the prediction accuracy.
[0018] That is, in the present invention, in the operation and inspection analysis method, the initial physical model and the initial behavioral model can be reduced in order through the singular value decomposition method and the principal component analysis method, thereby reducing the complexity of the model while maintaining a high prediction accuracy to obtain the corresponding physical model and the behavioral model.
[0019] Similarly, in the operation and inspection analysis method of the present invention, the computational complexity and storage requirements of the model are reduced through pruning and quantization techniques. The initial rule model is pruned through pruning technology to remove unimportant rule items to obtain a lightweight rule model; and the initial comprehensive state evaluation result obtained by subsequent fusion is quantized to obtain a lightweight comprehensive state evaluation result.
[0020] Accordingly, the present invention also discloses a computer medium, which specifically includes: a memory, a processor, and a computer program stored in the memory and executable on the processor; Wherein, the processor is used to implement the above-mentioned operation and inspection analysis method based on mechanism and data fusion drive of the present invention when executing the computer program.
[0021] In addition, the present invention also discloses an operation and inspection analysis system based on mechanism and data fusion drive, which includes: A model building module, which is used to build a digital geometric model, a physical model, a behavioral model and a rule model of the power equipment; A data acquisition module, which is used to collect real-time operating data of the power equipment; A data processing module, which is used to pre-process the real-time operation data and input the pre-processed real-time operation data into the digital geometry model, the physical model, the behavior model and the rule model to obtain the output of each model; A data fusion module, which fuses the outputs of each model based on a weighted fusion algorithm to obtain a comprehensive status assessment result of the power equipment; A prediction module predicts the future state change trend of the power equipment based on the comprehensive state evaluation result by using time series analysis and machine learning methods.
[0022] Furthermore, in the operation and inspection analysis system driven by mechanism and data fusion described in the present invention, the model construction module also includes: a reduction processing unit and a pruning processing unit to obtain the physical model and the behavior model after reduction, and obtain the rule model after pruning; the data fusion module also includes: a quantization processing unit to obtain the comprehensive state evaluation result after quantization processing.
[0023] The beneficial effects of the present invention are as follows: the operation and maintenance analysis method, medium and system based on mechanism and data fusion drive of the present invention realizes accurate assessment of the state of power transmission and transformation equipment and accurate prediction of future situation by constructing a high-approximation state assessment and situation prediction method driven by "model + data" fusion of digital twin geometric model, physical model, behavior model and rule model, and can improve the accuracy of component-level and equipment-level power operation and maintenance risk identification under complex working conditions. It has good promotion prospects and application value. BRIEF DESCRIPTION OF THE DRAWINGS
[0024] Figure 1 The present invention is a flowchart of the steps of the operation, inspection and analysis method based on mechanism and data fusion drive in one implementation mode.
[0025] Figure 2 It is a structural block diagram of the operation, inspection and analysis system based on mechanism and data fusion drive in one implementation mode of the present invention. DETAILED DESCRIPTION
[0026] In order to explain the technical content, achieved objectives and effects of the present invention in detail, the following is an explanation in combination with the implementation modes and the accompanying drawings.
[0027] Please refer to Figure 1 As shown, in this embodiment, the present invention discloses an operation and inspection analysis method based on mechanism and data fusion drive, which specifically includes the steps of: S1: Construct digital geometric model, physical model, behavioral model and rule model of power equipment; S2: collecting the real-time operation data of the power equipment and preprocessing it, so as to input the preprocessed real-time operation data into each model and obtain the output of each model; S3: using a weighted fusion algorithm to fuse the outputs of each model to obtain a comprehensive status assessment result of the power equipment; S4: Based on the comprehensive status assessment results, use time series analysis and machine learning methods to predict future status change trends of the power equipment.
[0028] That is to say, in this embodiment, the operation and inspection analysis method disclosed in the present invention can first construct a digital geometric model, a physical model, a behavioral model and a rule model of the power equipment to respectively describe the physical structure, multi-physical field behavior, dynamic behavior and empirical rules of the equipment; then, through data acquisition and preprocessing, the real-time operation data of the power equipment is collected and preprocessed, so that the preprocessed real-time operation data is input into each model; then, a weighted fusion algorithm is used to fuse the outputs of each model to obtain a comprehensive status assessment result of the equipment; finally, based on the fused status assessment result, time series analysis and machine learning methods are used to predict the future status change trend of the power equipment.
[0029] It should be noted that in the operation and inspection analysis method based on mechanism and data fusion drive of the present invention, in the above step S3, the above comprehensive state evaluation result is specifically: S = αMg + βMp + γMb + δMr Among them, S represents the comprehensive state assessment result; Mg, Mp, Mb, and Mr represent the outputs of the digital geometry model, physical model, behavior model, and rule model, respectively; α, β, γ, and δ represent different weight coefficients, respectively.
[0030] Accordingly, in the operation and inspection analysis method of the present invention, in the above step S4, a situation prediction model can be obtained by using time series analysis and machine learning methods to predict the future state change trend of the power equipment based on the situation prediction model; wherein the above situation prediction model is specifically: P(t)=f(S,t) Wherein, P(t) is the situation prediction model, f is the prediction function, S represents the comprehensive state evaluation result, t is the time variable, and f(S, t) represents the equipment state prediction of the power equipment at a specific time point t.
[0031] In addition, in the present invention, in order to improve the timeliness of prediction and computing efficiency, the operation and inspection analysis method based on mechanism and data fusion drive described in the present invention also applies model reduction and lightweight technology to solve the problem of long simulation and calculation time, thereby reducing latency and improving prediction timeliness while ensuring prediction accuracy.
[0032] Therefore, in the operation and inspection analysis method, in step S1, the initial physical model Mp′ and the initial behavior model Mb′ of the power equipment can be constructed first, and the initial physical model Mp′ and the initial behavior model Mb′ can be reduced in order by using the singular value decomposition method and the principal component analysis method, for example: Perform singular value decomposition on the initial physical model Mp′, retain the first k singular values, and obtain the physical model Mp after reduction; The initial behavior model Mb′ is subjected to principal component analysis, and the first i principal components are retained to obtain the reduced-order behavior model Mb.
[0033] Therefore, through the above-mentioned model reduction processing, the complexity of the model can be effectively reduced while maintaining a high prediction accuracy to obtain the final physical model Mp and behavioral model Mb.
[0034] Similarly, in the operation and inspection analysis method, pruning and quantization techniques are also used to reduce the computational complexity and storage requirements of the model; wherein, in step S1, the initial rule model Mr′ of the power equipment can be constructed first, and then the initial rule model Mr′ can be pruned through pruning technology to remove unimportant rule items to obtain a lightweight rule model Mr; and, in step S3, a weighted fusion algorithm can be used to fuse the outputs of each model to first obtain the initial comprehensive state assessment result S′ of the power equipment, and then the initial comprehensive state assessment result obtained by subsequent fusion is quantized to obtain a lightweight comprehensive state assessment result S.
[0035] Based on the above settings, the above-mentioned operation and inspection analysis method based on mechanism and data fusion driven by the present invention constructs a high-approximation state evaluation and situation prediction method driven by the "model + data" fusion of the digital twin geometric model, physical model, behavioral model and rule model, thereby effectively improving the precise evaluation of the state of substation equipment under complex working conditions and the accurate prediction of future situations, and studying model reduction and lightweight technology, while ensuring the prediction accuracy, effectively improving the prediction timeliness and reducing the digital thread delay.
[0036] Correspondingly, in the present invention, a computer medium is also disclosed, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor; wherein the processor can be used to implement the above-mentioned mechanism- and data fusion-driven operation and inspection analysis method of the present invention when executing the computer program.
[0037] Also, see Figure 2 As shown, in order to facilitate the execution of the above-mentioned operation and inspection analysis method of the present invention, the present invention also discloses an operation and inspection analysis system driven by mechanism and data fusion, which includes: a model building module, a data acquisition module, a data processing module, a data fusion module and a prediction module.
[0038] It should be noted that the above-mentioned model construction module is specifically used to construct digital geometric models, physical models, behavioral models and rule models of power equipment; wherein, in order to improve prediction timeliness and computational efficiency, the model construction module may specifically include: a reduction-order processing unit and a pruning processing unit, so as to utilize the above-mentioned reduction-order processing unit to process the reduced-order physical model and the behavioral model, and utilize the above-mentioned pruning processing unit to obtain the rule model after pruning.
[0039] Also, see Figure 2 As shown, the data acquisition module is used to collect real-time operation data of the power equipment; the data processing module is used to preprocess the real-time operation data, and input the preprocessed real-time operation data into the digital geometry model, the physical model, the behavior model and the rule model to obtain the output of each model; the data fusion module can fuse the output of each model based on the weighted fusion algorithm to obtain the comprehensive state evaluation result of the power equipment; the prediction module can receive the comprehensive state evaluation result, and based on the comprehensive state evaluation result, use time series analysis and machine learning methods to predict the future state change trend of the power equipment.
[0040] Accordingly, in practical applications, the comprehensive state evaluation results may also be stored. To improve storage efficiency, the data fusion module may further include a quantization processing unit to perform quantization to obtain the comprehensive state evaluation results after quantization.
[0041] From the above, it can be seen that the operation and inspection analysis method, medium and system based on mechanism and data fusion drive of the present invention realizes accurate assessment of the state of substation equipment and accurate prediction of future situation by constructing a high-approximation state assessment and situation prediction method driven by "model + data" fusion of digital twin geometric model, physical model, behavior model and rule model. It can accurately diagnose and predict and assist decision-making of power equipment under complex working conditions, and has good promotion prospects and application value.
[0042] Accordingly, in some preferred embodiments, the operation and inspection analysis method, medium and system also specifically incorporate model reduction and lightweight technology, which can effectively reduce the number of model parameters and computational complexity, and solve the problem of long simulation and calculation time, so as to effectively improve the prediction timeliness and reduce the digital thread delay on the basis of ensuring prediction accuracy.
[0043] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent transformations made using the contents of the present invention's specification and drawings, or directly or indirectly applied in related technical fields, are also included in the patent protection scope of the present invention.
Claims
1. A method for operation and inspection analysis based on mechanism and data fusion drive, characterized in that: Includes steps: S1: Construct digital geometric model, physical model, behavioral model and rule model of power equipment; S2: collecting the real-time operation data of the power equipment and preprocessing it, so as to input the preprocessed real-time operation data into each model and obtain the output of each model; S3: using a weighted fusion algorithm to fuse the outputs of each model to obtain a comprehensive status assessment result of the power equipment; S4: Based on the comprehensive status assessment results, use time series analysis and machine learning methods to predict future status change trends of the power equipment.
2. The operation inspection and analysis method based on mechanism and data fusion drive according to claim 1, characterized in that: In step S3, the comprehensive status evaluation result is specifically: S = αMg + βMp + γMb + δMr Among them, S represents the comprehensive state assessment result; Mg, Mp, Mb, and Mr represent the outputs of the digital geometry model, physical model, behavior model, and rule model, respectively; α, β, γ, and δ represent different weight coefficients, respectively.
3. The operation inspection and analysis method based on mechanism and data fusion drive according to claim 1 is characterized in that: In step S4, a situation prediction model is obtained by using time series analysis and machine learning methods to predict the future state change trend of the power equipment based on the situation prediction model. The situation prediction model is specifically: P(t)=f(S,t) Wherein, P(t) is the situation prediction model, f is the prediction function, S represents the comprehensive state evaluation result, t is the time variable, and f(S, t) represents the equipment state prediction of the power equipment at a specific time point t.
4. The operation and inspection analysis method based on mechanism and data fusion drive according to claim 1 is characterized in that: In step S1, an initial physical model and an initial behavioral model of the power equipment are constructed, and the initial physical model and the initial behavioral model are reduced in order by a singular value decomposition method and a principal component analysis method to obtain the physical model and the behavioral model.
5. The operation and inspection analysis method based on mechanism and data fusion drive according to claim 4 is characterized in that: In step S1, the initial physical model and the initial behavioral model are reduced in order, and the specific steps are as follows: Performing singular value decomposition on the initial physical model, retaining the first k singular values, and obtaining the physical model after reduction; Perform principal component analysis on the initial behavior model, retain the first i principal components, and obtain the reduced-order behavior model.
6. The operation and inspection analysis method based on mechanism and data fusion drive according to claim 1 is characterized in that: In step S1, an initial rule model of the electric power equipment is constructed, and the initial rule model is pruned by using pruning technology to remove unimportant rule items to obtain a lightweight rule model.
7. The operation and inspection analysis method based on mechanism and data fusion drive according to claim 1 is characterized in that: In step S3, a weighted fusion algorithm is used to fuse the outputs of each model to obtain an initial comprehensive state assessment result of the power equipment, the initial comprehensive state assessment result is quantized, and the floating-point parameters of the initial comprehensive state assessment result are converted into integers to obtain a lightweight comprehensive state assessment result.
8. A computer medium, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor; Wherein, the processor is used to implement an operation, inspection and analysis method based on mechanism and data fusion drive as described in any one of claims 1 to 7 when executing the computer program.
9. An operation and inspection analysis system based on mechanism and data fusion drive, characterized in that: include: A model building module, which is used to build a digital geometric model, a physical model, a behavioral model and a rule model of the power equipment; A data acquisition module, which is used to collect real-time operating data of the power equipment; A data processing module, which is used to pre-process the real-time operation data and input the pre-processed real-time operation data into the digital geometry model, the physical model, the behavior model and the rule model to obtain the output of each model; A data fusion module, which fuses the outputs of each model based on a weighted fusion algorithm to obtain a comprehensive status assessment result of the power equipment; A prediction module predicts the future state change trend of the power equipment based on the comprehensive state evaluation result by using time series analysis and machine learning methods.
10. The operation and inspection analysis system based on mechanism and data fusion drive according to claim 9, characterized in that: The model building module also includes: a reduction processing unit and a pruning processing unit to obtain the physical model and the behavior model after reduction, and obtain the rule model after pruning processing; the data fusion module also includes: a quantization processing unit to obtain the comprehensive state evaluation result after quantization processing.