Method, system and readable storage medium for dynamic settlement monitoring of power equipment

Abstract

The application provides a kind of dynamic settlement monitoring method, system and readable storage medium of electric power equipment, the method comprises: obtaining the environmental data and state data of each key point of electric power equipment, and constructs space-time coupling matrix;The feature extraction is carried out to the space-time coupling matrix after pre-processing, obtains a variety of settlement feature vectors, and all the variety of settlement feature vectors are fused to obtain fusion features;According to the fusion feature, a settlement prediction model is constructed, and a settlement prediction sequence is obtained according to the settlement prediction model, and the settlement prediction sequence contains the settlement prediction value of multiple future time points;The modal decomposition is carried out to the settlement prediction sequence, and the energy entropy of each IMF is calculated according to the modal decomposition result, and the mutation index of each point is calculated according to the energy entropy;According to the preset settlement early warning rule, the mutation index is monitored to obtain a monitoring result.The application can comprehensively, accurately and timely monitor the settlement of electric power equipment.

Description

Technical Field

[0001] This invention relates to the field of settlement monitoring technology, and in particular to a dynamic settlement monitoring method, system, and readable storage medium for power equipment. Background Technology

[0002] During the operation of power equipment, its settlement is affected by a variety of factors, and the settlement changes are dynamic and complex.

[0003] Traditional methods for monitoring the settlement of power equipment have significant limitations. Firstly, the data sources are often limited, focusing only on partial data about the equipment itself, such as displacement, while ignoring the influence of environmental factors and other critical data on settlement. This fails to comprehensively reflect the true settlement situation. Secondly, data processing and analysis methods are relatively outdated, lacking a systematic data processing workflow. This makes it difficult to accurately extract settlement-related features from massive amounts of data, resulting in low accuracy in settlement predictions. Consequently, it fails to detect sudden changes in equipment settlement risks in a timely manner, posing serious threats to the safe and stable operation of power equipment. Summary of the Invention

[0004] The purpose of this invention is to provide a dynamic settlement monitoring method, system, and readable storage medium for power equipment, aiming to solve the problem that traditional methods for monitoring the settlement of power equipment cannot detect sudden changes in equipment settlement in a timely manner, which poses a serious threat to the safe and stable operation of power equipment.

[0005] In a first aspect, the present invention provides a method for dynamic settlement monitoring of power equipment, the method comprising:

[0006] Acquire environmental and status data of key points of power equipment, and construct a spatiotemporal coupling matrix based on the environmental and status data;

[0007] The spatiotemporal coupling matrix is ​​preprocessed, and features are extracted from the preprocessed spatiotemporal coupling matrix to obtain multiple settlement feature vectors. All the multiple settlement feature vectors are then fused to obtain fused features.

[0008] A settlement prediction model is constructed based on the fusion features, and a settlement prediction sequence is obtained based on the settlement prediction model. The settlement prediction sequence contains settlement prediction values ​​for multiple future time points.

[0009] The settlement prediction sequence is subjected to mode decomposition, and the energy entropy of each IMF is calculated based on the mode decomposition results. The mutation index of each point is then calculated based on the energy entropy.

[0010] The mutation index is monitored according to the preset settlement early warning rules, and the monitoring results are obtained.

[0011] In some embodiments, the step of acquiring environmental data and status data of various key points of the power equipment, and constructing a spatiotemporal coupling matrix based on the environmental data and status data includes:

[0012] The environmental data includes temperature and humidity, and the status data includes liquid level displacement data, tilt angle data, strain data, and operating power.

[0013] Construct the spatiotemporal coupling matrix according to the following formula:

[0014] ;

[0015] in, Let be the spatiotemporal coupling matrix at time t. , These are the temperatures of the 1st and nth key points, respectively. , The humidity at the 1st and nth key points are respectively. , These are the liquid level displacement data for the 1st and nth key points, respectively. , The tilt angle data are for the 1st and nth key points, respectively. , These are the strain data for the 1st and nth key points, respectively. , These are the operating power of the 1st and nth key points, respectively.

[0016] In some embodiments, the step of preprocessing the spatiotemporal coupling matrix includes:

[0017] Calculate the mean and standard deviation of each type of data at time t based on the data in the same column of the spatiotemporal coupling matrix;

[0018] A mean matrix is ​​constructed based on the mean. The dimensions of the mean matrix, the standard deviation matrix, and the spatiotemporal coupling matrix are all the same. The values ​​in the i-th row and j-th column of the mean matrix and the standard deviation matrix are the mean and standard deviation of the corresponding data types, respectively.

[0019] The spatiotemporal coupling matrix is ​​preprocessed according to the following formula:

[0020] ;

[0021] in, This is the preprocessed spatiotemporal coupling matrix. Let be the mean matrix at time t. Let be the standard deviation matrix at time t. It is a positive integer. The enhancement factor ranges from (0, 0.5). This is for symbolic function operations.

[0022] In some embodiments, the step of extracting features from the preprocessed spatiotemporal coupling matrix to obtain multiple settlement feature vectors includes:

[0023] Temporal features are extracted from the preprocessed spatiotemporal coupling matrix according to the following formula:

[0024] ;

[0025] in, For the time-domain features of the j-th type of data at the i-th key point, , These represent the j-th type of data related to the i-th keypoint in the preprocessed spatiotemporal coupling matrix at time t. The eigenvalue corresponding to time step 1. The duration, including the total number of time points;

[0026] Spatial features are extracted from the preprocessed spatiotemporal coupling matrix according to the following formula:

[0027] ;

[0028] in, Let n be the spatial feature of the j-th type of data for the i-th keypoint, and n be the total number of keypoints. Let be the physical distance between the i-th keypoint and the k-th keypoint. For bandwidth parameters, The standardized absolute difference between the i-th keypoint and the j-th type of data of the k-th keypoint. It is the average of the standard deviations of the j-th type of data between the i-th keypoint and the k-th keypoint.

[0029] In some embodiments, the standardized absolute difference is calculated according to the following formula:

[0030] ;

[0031] in, The eigenvalues ​​of the j-th type of data corresponding to the k-th keypoint at time t in the preprocessed spatiotemporal coupling matrix are: , These are the median values ​​of all eigenvalues ​​corresponding to the j-th type of data at the i-th and k-th keypoints in the preprocessed spatiotemporal coupling matrix over the entire time period.

[0032] In some embodiments, the step of fusing all the multiple settlement feature vectors to obtain the fused feature includes:

[0033] All settlement feature vectors are fused according to the following formula:

[0034] ;

[0035] in, Let be the fusion feature of the i-th keypoint at time t. , These are the temporal and spatial characteristics of the j-th type of data at the i-th key point at time t, respectively.

[0036] The steps of constructing a settlement prediction model based on the fusion features and obtaining a settlement prediction sequence based on the settlement prediction model, wherein the settlement prediction sequence contains settlement prediction values ​​for multiple future time points, include:

[0037] A settlement prediction model is constructed based on the following formula:

[0038] ;

[0039] in, For settlement prediction sequences, This is the weight matrix. , These are the hidden states of the LSTM at time t and time t-1, respectively. For bias vectors, , , They are time t+1, time t+2, respectively. Predicted settlement values ​​at any given time.

[0040] In some embodiments, the steps of performing mode decomposition on the settlement prediction sequence, calculating the energy entropy of each IMF based on the mode decomposition results, and calculating the mutation index of each point based on the energy entropy include:

[0041] Modal decomposition is performed on the predicted settlement values ​​at various future times under the same key point to obtain K IMF components corresponding to each future time.

[0042] Divide the amplitude of the IMF component into M intervals and calculate the energy entropy of each IMF component:

[0043] ;

[0044] in, The energy entropy of the k-th IMF component. For the k-th IMF component at time... The energy probability under, , The k-th IMF component is at time [time]. ,time The amplitude below;

[0045] The mutation index is calculated using the following formula:

[0046] ;

[0047] in, The mutation index, The weight of the k-th IMF component.

[0048] In some embodiments, the step of monitoring the abrupt change index according to a preset settlement early warning rule and obtaining the monitoring result includes:

[0049] Check whether the mutation index of each key point is greater than the first preset threshold.

[0050] If the mutation index is greater than the first preset threshold, then the key point is a sedimentation mutation point, and an early warning message is issued at this time;

[0051] If the mutation index is less than or equal to the first preset threshold, then the key point is a normal point.

[0052] In a second aspect, the present invention provides a dynamic settlement monitoring system for power equipment, the system comprising:

[0053] The data acquisition module is used to acquire environmental and status data of various key points of the power equipment, and to construct a spatiotemporal coupling matrix based on the environmental and status data.

[0054] The feature extraction module is used to preprocess the spatiotemporal coupling matrix, extract features from the preprocessed spatiotemporal coupling matrix to obtain multiple settlement feature vectors, and fuse all the multiple settlement feature vectors to obtain fused features.

[0055] The model building module is used to build a settlement prediction model based on the fusion features and obtain a settlement prediction sequence based on the settlement prediction model. The settlement prediction sequence contains settlement prediction values ​​for multiple future time points.

[0056] The mutation index calculation module is used to perform mode decomposition on the settlement prediction sequence, calculate the energy entropy of each IMF based on the mode decomposition results, and calculate the mutation index of each point based on the energy entropy.

[0057] The monitoring module is used to monitor the mutation index according to the preset settlement early warning rules and obtain the monitoring results.

[0058] Thirdly, the present invention provides a readable storage medium that stores one or more programs that, when executed by a processor, implement the above-described dynamic settlement monitoring method for power equipment.

[0059] Fourthly, the present invention provides an electronic device, the electronic device comprising a memory and a processor, wherein:

[0060] The memory is used to store computer programs;

[0061] When the processor executes the computer program stored in the memory, it implements the above-mentioned dynamic settlement monitoring method for power equipment.

[0062] Compared with the prior art, the present invention has the following advantages:

[0063] This invention integrates multi-source data to construct a spatiotemporal coupling matrix. After preprocessing and multi-dimensional feature extraction and fusion, a precise settlement prediction model is constructed. Combined with modal decomposition and mutation index monitoring, it achieves comprehensive, accurate and timely monitoring of the dynamic settlement of power equipment. Specifically, it first acquires environmental and status data of each key point of the power equipment, and then constructs a spatiotemporal coupling matrix based on these multi-source data. This approach organically integrates data of different types and time periods, breaking the limitations of traditional methods that rely on a single data source. It comprehensively considers the impact of various factors on the settlement of power equipment, providing a rich and comprehensive data foundation for subsequent accurate analysis of settlement. Then, a mean matrix and standard deviation matrix are constructed, and the spatiotemporal coupling matrix is ​​standardized according to a specific formula. This eliminates the influence of dimensional differences and outliers between different types of data, making the data more standardized and improving its quality and usability. Temporal and spatial features are then extracted from the preprocessed spatiotemporal coupling matrix. By extracting these two different dimensions, the characteristics of power equipment settlement can be more comprehensively depicted. An LSTM model is then used to output a settlement prediction sequence containing settlement predictions for multiple future time points, achieving accurate prediction of future power equipment settlement. Finally, modal decomposition is performed on the settlement prediction sequence to obtain K IMF components corresponding to each future time point. The energy entropy of each IMF component is then calculated, and the mutation index at each point is calculated based on the energy entropy. The mutation index reflects the degree of abrupt change in equipment settlement, thereby enabling the monitoring of power equipment settlement. Attached Figure Description

[0064] Figure 1 This is a flowchart of a dynamic settlement monitoring method for power equipment according to an embodiment of the present invention;

[0065] Figure 2 This is a schematic diagram of the structure of a dynamic settlement monitoring system for power equipment proposed in an embodiment of the present invention.

[0066] The following detailed description, in conjunction with the accompanying drawings, will further illustrate the present invention. Detailed Implementation

[0067] 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. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Unless otherwise defined, the technical or scientific terms used herein should have the ordinary meaning understood by those skilled in the art. The terms "comprising" and similar expressions used herein mean that the element or object preceding the word covers the element or object listed after the word and its equivalents, but does not exclude other elements or objects.

[0068] like Figure 1 As shown, an embodiment of the present invention proposes a method for dynamic settlement monitoring of power equipment, the method comprising steps S101 to S105, wherein:

[0069] Step S101: Obtain environmental data and status data of each key point of the power equipment, and construct a spatiotemporal coupling matrix based on the environmental data and status data;

[0070] It should be noted that key points refer to representative locations on power equipment that are sensitive to equipment settlement and require special attention during monitoring, such as locations bearing heavy loads, the four corners of transformers, etc.

[0071] Furthermore, during the actual operation of power equipment, single data points are insufficient to accurately depict the complex and ever-changing settlement conditions. Environmental factors can affect the stability of the equipment foundation, and status data directly reflects the current settlement status of the equipment. Therefore, in order to achieve a comprehensive assessment of the settlement of power equipment, this embodiment requires the collection of both environmental and status data.

[0072] In some embodiments, the environmental data includes two types of data: temperature and humidity, and the state data includes four types of data: liquid level displacement data, tilt angle data, strain data, and operating power.

[0073] Furthermore, the environmental and state data are arranged and combined according to the time and key point dimensions to form a spatiotemporal coupling matrix containing three dimensions: time, key points, and data type. Specifically, the spatiotemporal coupling matrix is ​​constructed according to the following formula:

[0074] ;

[0075] in, Let be the spatiotemporal coupling matrix at time t. , These are the temperatures of the 1st and nth key points, respectively. , The humidity at the 1st and nth key points are respectively. , These are the liquid level displacement data for the 1st and nth key points, respectively. , The tilt angle data are for the 1st and nth key points, respectively. , These are the strain data for the 1st and nth key points, respectively. , These are the operating power of the 1st and nth key points, respectively.

[0076] In summary, the settlement of power equipment is a process that changes continuously over time and space. Data at different times reflects the development trend of settlement, data at different key points reflects the settlement situation of different parts of the equipment, and different types of data describe the relevant characteristics of settlement from different perspectives. For example, at different times of the day, the settlement of the equipment may vary due to changes in temperature and operating load; the settlement of key points at different locations of the equipment, such as the four corners and the center of the foundation, may also differ. By constructing a spatiotemporal coupling matrix and integrating data from these three dimensions, the dynamic changes in equipment settlement can be comprehensively and systematically reflected.

[0077] Step S102: Preprocess the spatiotemporal coupling matrix, extract features from the preprocessed spatiotemporal coupling matrix to obtain multiple settlement feature vectors, and fuse all the multiple settlement feature vectors to obtain fused features;

[0078] It should be noted that, since different types of data differ in their units and numerical ranges and may contain outliers, directly using the raw data will affect the accuracy of subsequent feature extraction and model construction. Preprocessing is required to eliminate these effects. Based on this, the mean and standard deviation of each type of data at time t are calculated first based on the data in the same column of the spatiotemporal coupling matrix.

[0079] A mean matrix is ​​constructed based on the mean. The dimensions of the mean matrix, the standard deviation matrix, and the spatiotemporal coupling matrix are all the same. The values ​​in the i-th row and j-th column of the mean matrix and the standard deviation matrix are the mean and standard deviation of the corresponding data types, respectively.

[0080] The spatiotemporal coupling matrix is ​​preprocessed according to the following formula:

[0081] ;

[0082] in, This is the preprocessed spatiotemporal coupling matrix. Let be the mean matrix at time t. Let be the standard deviation matrix at time t. It is a positive integer. The enhancement factor ranges from (0, 0.5). This is for symbolic function operations.

[0083] In summary, different types of data can vary significantly in terms of dimensions and numerical ranges. For example, temperature data might range from -10℃ to 50℃, while strain data might range at the micro-strain level. Directly using these raw data for subsequent analysis would lead to larger numerical values ​​dominating the analysis, while smaller values ​​would be neglected. By calculating the mean and standard deviation using the formulas described above and performing standardization, different types of data can be transformed into data with the same dimensions and similar numerical ranges, eliminating the influence of data differences and dimensions, and making the data comparable.

[0084] Furthermore, in some embodiments, temporal features are extracted from the preprocessed spatiotemporal coupling matrix according to the following formula:

[0085] ;

[0086] in, For the time-domain features of the j-th type of data at the i-th key point, , These represent the j-th type of data related to the i-th keypoint in the preprocessed spatiotemporal coupling matrix at time t. The eigenvalue corresponding to time step 1. The duration, including the total number of time points;

[0087] Spatial features are extracted from the preprocessed spatiotemporal coupling matrix according to the following formula:

[0088] ;

[0089] in, Let n be the spatial feature of the j-th type of data for the i-th keypoint, and n be the total number of keypoints. Let be the physical distance between the i-th keypoint and the k-th keypoint. For bandwidth parameters, The standardized absolute difference between the i-th keypoint and the j-th type of data of the k-th keypoint. It is the average of the standard deviations of the j-th type of data between the i-th keypoint and the k-th keypoint.

[0090] In summary, the formulas for obtaining time-domain characteristics described above, by averaging the absolute values ​​of the differences between data points at different times and then taking the logarithm, can effectively measure the magnitude of data variation over different time intervals. In power equipment settlement monitoring, this can be used to measure the changes in settlement-related data at different key points over time, revealing the average degree of change in settlement data at a particular key point over a period of time, thereby understanding the dynamic characteristics of equipment settlement.

[0091] Furthermore, in the calculation formula for obtaining spatial domain features, It functions similarly to weight adjustment. It dynamically adjusts the contribution of each element to the total based on physical distance. Setting the bandwidth parameter can control the sensitivity of this adjustment. In settlement monitoring scenarios, it rationally allocates the influence of different key points according to their correlation, allowing highly correlated data to play a greater role in the comprehensive measurement, making the obtained spatial features more reflective of key settlement information. Furthermore… Combining different types of characteristic values ​​helps to detect potential problems in the settlement of power equipment in a timely manner.

[0092] Furthermore, in some embodiments, the standardized absolute difference is calculated according to the following formula:

[0093] ;

[0094] in, The eigenvalues ​​of the j-th type of data corresponding to the k-th keypoint at time t in the preprocessed spatiotemporal coupling matrix are: , These are the median values ​​of all eigenvalues ​​corresponding to the j-th type of data at the i-th and k-th keypoints in the preprocessed spatiotemporal coupling matrix over the entire time period.

[0095] Furthermore, during the actual operation of power equipment, its settlement may be the result of the combined effects of uneven foundation settlement and changes in equipment operating load. In this case, considering only time-domain or spatial features is insufficient to accurately predict the equipment settlement. By fusing time-domain and spatial features, information from both time and space dimensions can be comprehensively considered, better adapting to complex settlement patterns. Based on this, all settlement feature vectors are fused according to the following formula:

[0096] ;

[0097] in, Let be the fusion feature of the i-th keypoint at time t. , These are the temporal and spatial characteristics of the j-th type of data at the i-th key point at time t.

[0098] Step S103: Construct a settlement prediction model based on the fusion features, and obtain a settlement prediction sequence based on the settlement prediction model. The settlement prediction sequence contains settlement prediction values ​​for multiple future time points.

[0099] It should be noted that in this step, the LSTM-based settlement prediction model can make full use of the information in the time series data to capture the long-term trend and short-term fluctuations of equipment settlement, thereby improving the accuracy and reliability of the prediction.

[0100] Specifically, a settlement prediction model is constructed based on the following formula:

[0101] ;

[0102] in, For settlement prediction sequences, This is the weight matrix. , These are the hidden states of the LSTM at time t and time t-1, respectively. For bias vectors, , , They are time t+1, time t+2, respectively. Predicted settlement values ​​at any given time.

[0103] Step S104: Perform mode decomposition on the settlement prediction sequence, calculate the energy entropy of each IMF based on the mode decomposition results, and calculate the mutation index of each point based on the energy entropy;

[0104] It should be noted that sudden settlement changes in power equipment can lead to serious consequences, such as equipment damage and power outages. Traditional monitoring methods often only detect obvious settlement changes that have already occurred, making it difficult to detect potential sudden risks in advance. By performing modal decomposition and energy entropy calculation on the settlement prediction sequence, we can further mine hidden information in the data, identify key points where settlement changes may occur in advance, and give maintenance personnel enough time to take measures to avoid accidents. Specifically, firstly, modal decomposition is performed on the settlement prediction values ​​at various future times under the same key point to obtain K IMF components corresponding to each future time.

[0105] Divide the amplitude of the IMF component into M intervals and calculate the energy entropy of each IMF component:

[0106] ;

[0107] in, The energy entropy of the k-th IMF component. For the k-th IMF component at time... The energy probability under, , The k-th IMF component is at time [time]. ,time The amplitude below;

[0108] The mutation index is calculated using the following formula:

[0109] ;

[0110] in, The mutation index, The weight of the k-th IMF component.

[0111] Step S105: Monitor the mutation index according to the preset settlement early warning rules and obtain the monitoring results.

[0112] In this step, the preset settlement early warning rule is as follows:

[0113] Check whether the mutation index of each key point is greater than the first preset threshold.

[0114] If the mutation index is greater than the first preset threshold, the key point is a settlement mutation point, which means that a large settlement mutation will occur at the key point. At this time, an early warning message is issued to inform and arrange personnel to reinforce the part in advance to prevent the equipment from tilting or collapsing.

[0115] If the mutation index is less than or equal to the first preset threshold, then the key point is a normal point.

[0116] In summary, based on the above-mentioned dynamic settlement monitoring method for power equipment, a spatiotemporal coupling matrix is ​​constructed by integrating multi-source data. After preprocessing and multi-dimensional feature extraction and fusion, an accurate settlement prediction model is constructed. Combined with modal decomposition and catastrophe index monitoring, comprehensive, accurate and timely monitoring of dynamic settlement of power equipment is achieved. Specifically, environmental and status data of each key point of the power equipment are first acquired, and a spatiotemporal coupling matrix is ​​constructed based on these multi-source data. This approach organically integrates data of different types and time periods, breaking the limitations of traditional methods that rely on a single data source. It comprehensively considers the impact of various factors on the settlement of power equipment, providing a rich and comprehensive data foundation for subsequent accurate analysis of settlement. Then, a mean matrix and standard deviation matrix are constructed, and the spatiotemporal coupling matrix is ​​standardized according to a specific formula. This eliminates the influence of dimensional differences and outliers between different types of data, making the data more standardized and improving its quality and usability. Temporal and spatial features are then extracted from the preprocessed spatiotemporal coupling matrix. By extracting these two different dimensions, the characteristics of power equipment settlement can be more comprehensively depicted. An LSTM model is then used to output a settlement prediction sequence containing settlement predictions for multiple future time points, achieving accurate prediction of future power equipment settlement. Finally, modal decomposition is performed on the settlement prediction sequence to obtain K IMF components corresponding to each future time point. The energy entropy of each IMF component is then calculated, and the mutation index at each point is calculated based on the energy entropy. The mutation index reflects the degree of abrupt change in equipment settlement, thereby enabling the monitoring of power equipment settlement.

[0117] like Figure 2 As shown in the figure, an embodiment of the present invention provides a dynamic settlement monitoring system for power equipment, the system comprising:

[0118] The data acquisition module 10 is used to acquire environmental data and status data of various key points of the power equipment, and to construct a spatiotemporal coupling matrix based on the environmental data and status data.

[0119] The feature extraction module 20 is used to preprocess the spatiotemporal coupling matrix, extract features from the preprocessed spatiotemporal coupling matrix to obtain multiple settlement feature vectors, and fuse all the multiple settlement feature vectors to obtain fused features;

[0120] The model building module 30 is used to build a settlement prediction model based on the fusion features and obtain a settlement prediction sequence based on the settlement prediction model. The settlement prediction sequence contains settlement prediction values ​​for multiple future time points.

[0121] The mutation index calculation module 40 is used to perform mode decomposition on the settlement prediction sequence, calculate the energy entropy of each IMF based on the mode decomposition results, and calculate the mutation index of each point based on the energy entropy.

[0122] The monitoring module 50 is used to monitor the mutation index according to the preset settlement early warning rules and obtain the monitoring results.

[0123] In another aspect, the present invention also proposes a readable storage medium having stored one or more programs thereon, which, when executed by a processor, implement the above-described method for monitoring the dynamic settlement of power equipment.

[0124] In another aspect, the present invention also proposes an electronic device, including a memory and a processor, wherein the memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory to realize the above-mentioned dynamic settlement monitoring method for power equipment.

[0125] Those skilled in the art will understand that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can mean any means that can contain stored, communicated, propagated, or transmitted programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0126] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0127] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0128] While embodiments of the present invention have been described in detail above, it will be apparent to those skilled in the art that various modifications and variations can be made to these embodiments. However, it should be understood that such modifications and variations fall within the scope and spirit of the invention as set forth in the claims. Furthermore, the invention described herein may have other embodiments and can be implemented or carried out in various ways.

Claims

1. A method for monitoring the dynamic settlement of power equipment, characterized in that, The method includes: Acquire environmental and status data of key points of power equipment, and construct a spatiotemporal coupling matrix based on the environmental and status data; The spatiotemporal coupling matrix is ​​preprocessed, and features are extracted from the preprocessed spatiotemporal coupling matrix to obtain multiple settlement feature vectors. All the multiple settlement feature vectors are then fused to obtain fused features. A settlement prediction model is constructed based on the fusion features, and a settlement prediction sequence is obtained based on the settlement prediction model. The settlement prediction sequence contains settlement prediction values ​​for multiple future time points. The settlement prediction sequence is subjected to mode decomposition, and the energy entropy of each IMF is calculated based on the mode decomposition results. The mutation index of each point is then calculated based on the energy entropy. Modal decomposition is performed on the predicted settlement values ​​at various future times under the same key point to obtain K IMF components corresponding to each future time. Divide the amplitude of the IMF component into M intervals and calculate the energy entropy of each IMF component: ； in, The energy entropy of the k-th IMF component. For the k-th IMF component at time... The energy probability under, , The k-th IMF component is at time [time]. ,time The amplitude below; The mutation index is calculated using the following formula: ； in, The mutation index, The weight of the k-th IMF component; The mutation index is monitored according to the preset settlement early warning rules, and the monitoring results are obtained.

2. The method for dynamic settlement monitoring of power equipment according to claim 1, characterized in that, The steps of acquiring environmental and status data of various key points of the power equipment and constructing a spatiotemporal coupling matrix based on the environmental and status data include: The environmental data includes temperature and humidity, and the status data includes liquid level displacement data, tilt angle data, strain data, and operating power. Construct the spatiotemporal coupling matrix according to the following formula: ； in, Let be the spatiotemporal coupling matrix at time t. , These are the temperatures of the 1st and nth key points, respectively. , The humidity at the 1st and nth key points are respectively. , These are the liquid level displacement data for the 1st and nth key points, respectively. , The tilt angle data are for the 1st and nth key points, respectively. , These are the strain data for the 1st and nth key points, respectively. , These are the operating power of the 1st and nth key points, respectively.

3. The method for dynamic settlement monitoring of power equipment according to claim 2, characterized in that, The preprocessing steps for the spatiotemporal coupling matrix include: Calculate the mean and standard deviation of each type of data at time t based on the data in the same column of the spatiotemporal coupling matrix; A mean matrix is ​​constructed based on the mean, and a standard deviation matrix is ​​constructed based on the standard deviation. The dimensions of the mean matrix, the standard deviation matrix, and the spatiotemporal coupling matrix are all identical. The values ​​in the i-th row and j-th column of the mean matrix and the standard deviation matrix are the mean and standard deviation of the corresponding data types, respectively. The spatiotemporal coupling matrix is ​​preprocessed according to the following formula: ； in, This is the preprocessed spatiotemporal coupling matrix. Let be the mean matrix at time t. Let be the standard deviation matrix at time t. It is a positive integer. The enhancement factor ranges from (0, 0.5). This is for symbolic function operations.

4. The dynamic settlement monitoring method for power equipment according to claim 3, characterized in that, The step of extracting features from the preprocessed spatiotemporal coupling matrix to obtain various settlement feature vectors includes: Temporal features are extracted from the preprocessed spatiotemporal coupling matrix according to the following formula: ； in, For the time-domain features of the j-th type of data at the i-th key point, , These represent the j-th type of data related to the i-th keypoint in the preprocessed spatiotemporal coupling matrix at time t. The eigenvalue corresponding to time step 1. The duration, including the total number of time points; Spatial features are extracted from the preprocessed spatiotemporal coupling matrix according to the following formula: ； in, Let n be the spatial feature of the j-th type of data for the i-th keypoint, and n be the total number of keypoints. Let be the physical distance between the i-th keypoint and the k-th keypoint. For bandwidth parameters, The standardized absolute difference between the i-th keypoint and the j-th type of data of the k-th keypoint. It is the average of the standard deviations of the j-th type of data between the i-th keypoint and the k-th keypoint.

5. The dynamic settlement monitoring method for power equipment according to claim 4, characterized in that, The standardized absolute difference is calculated using the following formula: ； in, The eigenvalues ​​of the j-th type of data corresponding to the k-th keypoint at time t in the preprocessed spatiotemporal coupling matrix are: , These are the median values ​​of all eigenvalues ​​corresponding to the j-th type of data at the i-th and k-th keypoints in the preprocessed spatiotemporal coupling matrix over the entire time period.

6. The dynamic settlement monitoring method for power equipment according to claim 4, characterized in that, The step of fusing all the multiple settlement feature vectors to obtain the fused features includes: All settlement feature vectors are fused according to the following formula: ； in, Let be the fusion feature of the i-th keypoint at time t. , These are the temporal and spatial characteristics of the j-th type of data at the i-th key point at time t, respectively. The steps of constructing a settlement prediction model based on the fusion features and obtaining a settlement prediction sequence based on the settlement prediction model, wherein the settlement prediction sequence contains settlement prediction values ​​for multiple future time points, include: A settlement prediction model is constructed based on the following formula: ； in, For settlement prediction sequences, This is the weight matrix. , These are the hidden states of the LSTM at time t and time t-1, respectively. For bias vectors, , , They are time t+1, time t+2, respectively. Predicted settlement values ​​at any given time.

7. The method for dynamic settlement monitoring of power equipment according to claim 6, characterized in that, The step of monitoring the abrupt change index according to a preset settlement early warning rule and obtaining the monitoring results includes: Check whether the mutation index of each key point is greater than the first preset threshold. If the mutation index is greater than the first preset threshold, then the key point is a sedimentation mutation point, and an early warning message is issued at this time; If the mutation index is less than or equal to the first preset threshold, then the key point is a normal point.

8. A dynamic settlement monitoring system for power equipment, characterized in that, The system includes: The data acquisition module is used to acquire environmental and status data of various key points of the power equipment, and to construct a spatiotemporal coupling matrix based on the environmental and status data. The feature extraction module is used to preprocess the spatiotemporal coupling matrix, extract features from the preprocessed spatiotemporal coupling matrix to obtain multiple settlement feature vectors, and fuse all the multiple settlement feature vectors to obtain fused features. The model building module is used to build a settlement prediction model based on the fusion features and obtain a settlement prediction sequence based on the settlement prediction model. The settlement prediction sequence contains settlement prediction values ​​for multiple future time points. The mutation index calculation module is used to perform mode decomposition on the settlement prediction sequence, calculate the energy entropy of each IMF based on the mode decomposition results, and calculate the mutation index of each point based on the energy entropy. Modal decomposition is performed on the predicted settlement values ​​at various future times under the same key point to obtain K IMF components corresponding to each future time. Divide the amplitude of the IMF component into M intervals and calculate the energy entropy of each IMF component: ； in, The energy entropy of the k-th IMF component. For the k-th IMF component at time... The energy probability under, , The k-th IMF component is at time [time]. ,time The amplitude below; The mutation index is calculated using the following formula: ； in, The mutation index, The weight of the k-th IMF component; The monitoring module is used to monitor the mutation index according to the preset settlement early warning rules and obtain the monitoring results.

9. A readable storage medium, characterized in that, The readable storage medium stores one or more programs that, when executed by a processor, implement the dynamic settlement monitoring method for power equipment as described in any one of claims 1-7.

Images