Full-life-cycle intelligent monitoring and health assessment method and system for substation equipment

Abstract

The invention relates to a full-life-cycle intelligent monitoring and health assessment method and system for substation equipment, and belongs to the technical field of substation equipment monitoring. Therefore, the health degree information of each piece of substation equipment is obtained through the life cycle monitoring model of the substation equipment, early warning is carried out according to the health degree information of each piece of substation equipment, early warning information is generated, and finally a related maintenance scheme is generated based on the early warning information in combination with the aging evolution factor data in the target area. According to the method, the data of the collection equipment is processed through federated learning, the size of the collected data is dynamically adjusted by combining the state and operational capability of the actual equipment, and the potential contribution of local data distribution of the equipment to the improvement of the current global model is comprehensively evaluated, so that the deviation of model aggregation of the federated learning is reduced; the prediction accuracy of the life cycle monitoring model is improved, and the reliability of early warning is improved.

Description

Technical Field

[0001] This invention relates to the field of technology, and in particular to a method and system for intelligent monitoring and health assessment of substation equipment throughout its entire life cycle. Background Technology

[0002] Substations are a crucial component of power systems, with primary functions including voltage transformation, receiving and distributing electrical energy, controlling power flow, and voltage regulation. Substations connect power grids of various voltage levels through internal transformers, ensuring efficient power transmission and distribution. Substations require various types of equipment, including: transformers, switches, small transformers, reactive power devices, and other auxiliary equipment. Transformers are core equipment used to change voltage levels. Switches include circuit breakers and disconnectors, used for circuit opening, closing, and protection. Small transformers typically refer to surge arresters, voltage transformers, current transformers, and capacitors. Reactive power devices compensate for reactive power and improve the power factor. Other equipment and auxiliary devices include wave traps, insulators, high-voltage bushings, conductors, grounding devices, secondary equipment, and high-voltage DC equipment. Substation equipment typically operates in heterogeneous environments. Differences in geographical location, operating load, and aging stage can lead to varying degrees of degradation. Under Non-IID data, federated learning models are prone to bias, resulting in inaccurate predictions from lifecycle monitoring models. This directly impacts the reliability of early warnings and may render generated maintenance plans ineffective. If the lifecycle monitoring model fails due to data issues, the entire early warning and maintenance chain will collapse, potentially leading to undetected equipment failures, increased maintenance costs, and safety risks. Summary of the Invention

[0003] This invention overcomes the shortcomings of existing technologies and provides a method and system for intelligent monitoring and health assessment of substation equipment throughout its entire life cycle.

[0004] To achieve the above objectives, the technical solution adopted by the present invention is as follows: The first aspect of this invention provides a method for intelligent monitoring and health assessment of substation equipment throughout its entire lifecycle, comprising the following steps: Data acquisition devices are set up at the equipment end to collect aging evolution factor data and performance characteristic data of substation equipment, and the data is processed based on federated learning. Collect aging evolution factor data and substation equipment performance characteristic data after processing by federated learning, and construct a life cycle monitoring model for substation equipment. The health information of each substation device is obtained through the life cycle monitoring model of the substation equipment, and early warning information is generated based on the health information of each substation device. Based on the early warning information and the aging evolution factor data in the target area, a relevant maintenance plan is generated.

[0005] Furthermore, in the aforementioned intelligent monitoring and health assessment method for the entire lifecycle of substation equipment, data acquisition equipment is installed at the equipment end to collect aging evolution factor data and performance characteristic data of the substation equipment, specifically including: Configure data acquisition nodes on the local end, collect data samples from the data acquisition nodes, obtain the computing power data of each data acquisition node, and initialize the size information of the collected data according to the computing power data of the data acquisition nodes. Data is collected through data acquisition nodes based on the size information of the collected data. The collected data is then subjected to knowledge distillation and input into the classroom network to obtain knowledge feature representations. Obtain the data size information of the knowledge feature representation, and determine whether the data size information of the knowledge feature representation is greater than the computing power data of the data acquisition node; When the data size information represented by the knowledge feature is greater than the computing power data of the data acquisition node, the size information of the acquired data is reduced until it is no greater than the computing power data of the data acquisition node, and the data is stored at the local end of the substation equipment.

[0006] Furthermore, in the aforementioned intelligent monitoring and health assessment method for the entire lifecycle of substation equipment, data processing is based on federated learning, specifically including: A federated learning model is constructed to obtain the distribution information and data update frequency of local data of substation equipment, and importance evaluation data features are set according to the distribution information and data update frequency of the local data of substation equipment. The federated learning model is used to learn the data of each substation. The quantization bit width of each layer or parameter is dynamically adjusted according to the statistical characteristics of the model update. A lightweight autoencoder is used to update and compress the high-dimensional data into a low-dimensional vector. After aggregating the low-dimensional representation, it is decoded back to the model for updating and deployed on the server as part of the global model. The local computational load and communication frequency of each device are dynamically adjusted according to the status of the substation equipment, the importance of the data, and the current model status. The potential contribution of the local data distribution of the equipment to the improvement of the current global model is evaluated, and substation equipment with a contribution value higher than the contribution value threshold is scheduled for local training. Different local iteration numbers are allocated according to the status of the substation equipment.

[0007] Furthermore, in the aforementioned intelligent monitoring and health assessment method for the entire lifecycle of substation equipment, aging evolution factor data processed through federated learning and performance characteristic data of substation equipment are collected, and a lifecycle monitoring model for substation equipment is constructed, specifically as follows: The data on aging evolution factors processed by federated learning and the performance characteristics data of substation equipment are collected. The substation equipment is treated as a node, the electrical connection as an edge, and the aging evolution factor data as a graph attribute. Historical fault data is collected based on the performance characteristic data of the substation equipment, the fault types of the historical fault data are obtained, and the impact weights on the lifespan of the substation equipment are dynamically adjusted based on the fault types. The nodes, edges, and graph attributes are connected to construct a topology graph. A lifecycle monitoring model for substation equipment is built based on a deep learning network. The topology graph is then input into the lifecycle monitoring model for substation equipment for training. The lifespan status of substation equipment is used as the output of the model, and the model parameters are continuously updated. When the prediction accuracy of the model reaches the preset prediction accuracy, the model parameters of the lifespan monitoring model of substation equipment are saved.

[0008] Furthermore, in the aforementioned intelligent monitoring and health assessment method for the entire lifecycle of substation equipment, the health information of each piece of substation equipment is obtained through the lifecycle monitoring model of the substation equipment, specifically as follows: The aging evolution factor data and performance characteristic data of the substation equipment within a preset time are obtained and input into the life cycle monitoring model of the substation equipment for prediction, so as to obtain the life cycle status of each substation equipment. The health information of each substation device is determined based on its lifespan status, and the health information of each substation device is output.

[0009] Furthermore, in the aforementioned intelligent monitoring and health assessment method for the entire lifecycle of substation equipment, early warnings are generated based on the health information of each substation device, specifically including: Set health evaluation index data, and determine whether the health information of each substation device is greater than the health evaluation index data; When the health information of the substation equipment is greater than the health evaluation index data, the distribution information of the substation equipment corresponding to each health information greater than the health evaluation index data is obtained. Early warning information is generated based on the distribution information of substation equipment corresponding to each health status information that is greater than the health status evaluation index data.

[0010] Furthermore, in the aforementioned intelligent monitoring and health assessment method for the entire lifecycle of substation equipment, a relevant maintenance plan is generated based on the early warning information combined with aging evolution factor data in the target area, specifically including: The damage probability value of each substation equipment that generates early warning information is obtained under the aging evolution factor data of the target area. The cost data information of each damaged equipment is obtained, and the total damage cost data is calculated based on the cost data information of each damaged equipment and the damage probability value. The replacement plan for the damaged equipment is initialized based on the total damage cost data. The replacement plan for the damaged equipment is used as a variable and introduced into a genetic algorithm. The total maintenance cost data of the variable is calculated and then input into the genetic algorithm for inheritance. Determine whether the total maintenance cost data of the variable is greater than the preset total maintenance cost threshold data. If the total maintenance cost data of the variable is not greater than the preset total maintenance cost threshold data, output the current variable. If the total cost of the maintenance plan for the variable is greater than the preset total cost threshold for the maintenance plan, the variable will continue to be inherited until it is no greater than the preset total cost threshold for the maintenance plan.

[0011] The second aspect of the present invention provides a substation equipment full life cycle intelligent monitoring and health assessment system, including a memory and a processor. The memory includes a substation equipment full life cycle intelligent monitoring and health assessment method program. When the substation equipment full life cycle intelligent monitoring and health assessment method program is executed by the processor, it implements the steps of the substation equipment full life cycle intelligent monitoring and health assessment method as described in any one of the claims.

[0012] A third aspect of this invention provides an intelligent monitoring and health assessment device for the entire lifecycle of substation equipment, comprising: The data processing module is used to set up data acquisition devices on the equipment side, collect aging evolution factor data and performance characteristic data of substation equipment through the data acquisition devices, and process the data based on federated learning. The model building module is used to collect aging evolution factor data and substation equipment performance characteristic data after being processed by federated learning, and to build a life cycle monitoring model for substation equipment. The early warning module is used to obtain the health information of each substation device through the life cycle monitoring model of the substation equipment, and to issue early warnings based on the health information of each substation device, thereby generating early warning information. The maintenance plan generation module is used to generate relevant maintenance plans based on the early warning information and the aging evolution factor data in the target area.

[0013] The fourth aspect of the present invention provides a computer-readable storage medium, including a method program for intelligent monitoring and health assessment of the entire life cycle of substation equipment. When the method program is executed by a processor, it implements the steps of the method for intelligent monitoring and health assessment of the entire life cycle of substation equipment as described in any one of the claims.

[0014] This invention addresses the shortcomings of the prior art and has the following beneficial effects: This invention involves setting up data acquisition devices at the equipment end to collect aging evolution factor data and performance characteristic data of substation equipment. The data is processed using federated learning to collect the processed aging evolution factor data and performance characteristic data, and a lifecycle monitoring model for the substation equipment is constructed. This model obtains the health information of each piece of equipment and generates early warnings based on the health information. Finally, based on the early warnings and aging evolution factor data in the target area, relevant maintenance plans are generated. This invention uses federated learning to process the data from the acquisition devices, dynamically adjusting the data acquisition size based on the actual equipment status and computing power. Furthermore, it comprehensively evaluates the potential contribution of local data distribution to the improvement of the current global model, reducing biases in federated learning model aggregation, improving the predictive accuracy of the lifecycle monitoring model, and enhancing the reliability of early warnings. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other embodiments can be obtained from these drawings without creative effort.

[0016] Figure 1 This diagram illustrates the overall flowchart of the intelligent monitoring and health assessment method for the entire lifecycle of substation equipment. Figure 2 The system block diagram of the intelligent monitoring and health assessment system for the entire life cycle of substation equipment is shown. Figure 3 A schematic diagram of a substation equipment intelligent monitoring and health assessment device covering the entire lifecycle of substation equipment is shown. Detailed Implementation

[0017] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0018] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0019] like Figure 1 As shown, the first aspect of this invention provides a method for intelligent monitoring and health assessment of substation equipment throughout its entire lifecycle, comprising the following steps: Data acquisition devices are set up at the equipment end to collect aging evolution factor data and performance characteristic data of substation equipment, and the data is processed based on federated learning. Collect aging evolution factor data and substation equipment performance characteristic data after processing by federated learning, and construct a life cycle monitoring model for substation equipment. The health information of each substation device is obtained through the life cycle monitoring model of substation equipment, and early warning information is generated based on the health information of each substation device. Based on early warning information and aging evolution factor data in the target area, relevant maintenance plans are generated.

[0020] It should be noted that this invention processes data from the acquisition devices using federated learning, dynamically adjusting the data acquisition scale based on the actual device status and computing power. Furthermore, it comprehensively evaluates the potential contribution of local data distribution to the improvement of the current global model, thereby reducing biases in federated learning model aggregation, improving the predictive accuracy of the lifecycle monitoring model, and enhancing the reliability of early warnings. Aging evolution factor data includes temperature, humidity, and other data.

[0021] Furthermore, in the intelligent monitoring and health assessment method for the entire life cycle of substation equipment, data acquisition devices are installed at the equipment end to collect aging evolution factor data and performance characteristic data of the substation equipment, specifically including: Configure data acquisition nodes on the local end, collect data samples from the data acquisition nodes, obtain the computing power data of each data acquisition node, and initialize the size information of the collected data based on the computing power data of the data acquisition nodes. The computing power data includes the amount of data processed per unit time, the data processing rate, etc. Each data acquisition node has a certain computing power, and the units of data size information include Kb, Gb, Tb, etc.

[0022] Data is collected through data acquisition nodes based on the size of the collected data. The collected data is then subjected to knowledge distillation and input into the classroom network to obtain knowledge feature representations. It's important to note that during local training, the device not only trains its model on its own data but also learns from a "teacher-hinted model" provided by the knowledge distillation server. This model represents the output of the global model on a public proxy dataset or a generalized set of features. After local training, the device transmits the difference between its local model and the "teacher-hinted model"—the difference being the knowledge distillation loss gradient or the refined feature differences—as part or a major component of the update to the server. This leverages knowledge distillation to force the local model to learn more general knowledge consistent with the global objective, resulting in a more compact and robust learning profile than the original gradient.

[0023] Obtain the data size information of the knowledge feature representation, and determine whether the data size information of the knowledge feature representation is greater than the computing power data of the data acquisition node; When the size of the knowledge feature data is greater than the computing power of the data acquisition node, the size of the acquired data is reduced until it is no greater than the computing power of the data acquisition node, and the data is stored at the local end of the substation equipment.

[0024] It should be noted that when the size of the knowledge feature representation data is greater than the computing power of the data acquisition node, the size of the acquired data is reduced. This fully considers that the data after knowledge distillation is not greater than the computing power of the data acquisition node, thus improving the rationality of updating the model.

[0025] Furthermore, in the intelligent monitoring and health assessment method for the entire lifecycle of substation equipment, data processing is based on federated learning, specifically including: Construct a federated learning model to obtain the distribution information and update frequency of local data of substation equipment, and set importance evaluation data features based on the distribution information and update frequency of local data of substation equipment. It should be noted that federated learning is a technology that enables multiple data owners to collaboratively train and share machine learning models while ensuring that local training data is not made public. The distribution information of local data for substation equipment includes the distribution location of substation equipment, such as geographical location, and the importance evaluation data characteristics include severe importance, moderate importance, high importance, and low importance.

[0026] The method for setting importance evaluation data characteristics based on the distribution information of local data of substation equipment and the number of data updates specifically includes: The distribution information of the local data of the substation equipment is statistically analyzed based on historical geological disaster event data, and the damage probability value under the distribution information of the local data of the substation equipment is calculated. Set a damage probability threshold evaluation index to determine whether the damage probability value under the historical geological disaster event data information of the distribution information of the local data of the substation equipment is greater than the damage probability threshold evaluation index; When the damage probability value under the historical geological disaster event data information of the distribution information of the local data of the substation equipment is greater than the damage probability threshold evaluation index, the number of data updates for the corresponding substation equipment is increased, and the highest level of importance evaluation data feature is configured. When the probability of damage under the historical geological disaster event data information of the distribution information of the local data of the substation equipment is not greater than the damage probability threshold evaluation index, the number of times the data of the current substation equipment is updated is maintained.

[0027] It should be noted that this method can further optimize the use of higher precision for layers with drastic changes and lower precision for relatively stable layers. Combined with a non-uniform quantization codebook, it can make more effective use of finite bits.

[0028] The federated learning model learns from the data of each substation, dynamically adjusts the quantization bit width of each layer or parameter based on the statistical characteristics of the model update, and uses a lightweight autoencoder to update and compress high-dimensional data into a low-dimensional vector. It should be noted that, during the learning process, federated learning models update new features and statistically analyze these updates to dynamically adjust the quantization bit width of each layer or parameter. Higher precision is used for layers or parameters with more frequent changes, while lower precision is used for relatively stable ones. This, combined with a non-uniform quantization codebook, allows for more effective use of limited bits. Lightweight autoencoders include TAESD lightweight autoencoders and MAE lightweight autoencoders, among others.

[0029] After aggregating the low-dimensional representation, it is decoded back to the model for updating and deployed on the server as part of the global model. The local computational load and communication frequency of each device are dynamically adjusted according to the status of the substation equipment, the importance of the data, and the current model status. It should be noted that the status of substation equipment includes numerical control parameters such as power consumption, computing power, and network bandwidth.

[0030] The potential contribution of the local data distribution of the equipment to the improvement of the current global model is evaluated, and substation equipment with a contribution value higher than the contribution value threshold is scheduled for local training. Different local iteration numbers are allocated according to the status of the substation equipment.

[0031] It should be noted that this method can break the pattern of fixed local iterations, achieve global optimization of resource utilization and model performance, avoid wasting resources on resource-constrained devices, comprehensively evaluate the potential contribution of device local data distribution to the current global model improvement, reduce the bias in federated learning model aggregation, improve the prediction accuracy of lifecycle monitoring models, and enhance the reliability of early warning.

[0032] Furthermore, in the intelligent monitoring and health assessment method for the entire life cycle of substation equipment, aging evolution factor data processed through federated learning and performance characteristic data of substation equipment are collected, and a life cycle monitoring model for substation equipment is constructed, specifically as follows: The data on aging evolution factors processed by federated learning and the performance characteristics data of substation equipment are collected. The substation equipment is treated as a node, the electrical connection as an edge, and the aging evolution factor data as a graph attribute. Historical fault data is collected based on the performance characteristics of substation equipment, the fault types of historical fault data are obtained, and the impact weights on the lifespan of substation equipment are dynamically adjusted based on the fault types. It should be noted that fault types include single-point faults, residual faults, two-point faults, multi-point faults, latent faults, electrical faults, mechanical faults, and control faults. For example, the weight of a single-point fault is set to 0.1, and the weight of a multi-point fault is set to 0.7. Connect nodes, edges, and graph attributes to construct a topology graph. Build a lifecycle monitoring model for substation equipment based on a deep learning network. Input the topology graph into the lifecycle monitoring model for substation equipment for training. The lifespan status of substation equipment is used as the output of the model, and the model parameters are continuously updated. When the prediction accuracy of the model reaches the preset prediction accuracy, the model parameters of the lifespan monitoring model of substation equipment are saved.

[0033] It should be noted that this method can be used to create a life cycle monitoring model for substation equipment, thereby predicting the lifespan of the equipment.

[0034] Furthermore, in the intelligent monitoring and health assessment method for the entire life cycle of substation equipment, the health information of each piece of substation equipment is obtained through the life cycle monitoring model of substation equipment, specifically: The aging evolution factor data and performance characteristic data of substation equipment within a preset time are obtained and input into the life cycle monitoring model of substation equipment for prediction, so as to obtain the life status of each substation equipment. The health information of each substation device is determined based on its lifespan status, and the health information of each substation device is output.

[0035] It should be noted that the lifespan status of substation equipment can be divided into multiple categories, and different lifespan statuses correspond to different levels of health. For example, a lifespan status of 100 days corresponds to a low level of health.

[0036] Furthermore, in the intelligent monitoring and health assessment method for the entire lifecycle of substation equipment, early warnings are generated based on the health information of each piece of substation equipment. Specifically, these include: Set health evaluation index data and determine whether the health information of each substation device is greater than the health evaluation index data. When the health information of substation equipment is greater than the health evaluation index data, the distribution information of the substation equipment corresponding to each health information greater than the health evaluation index data is obtained. Early warning information is generated based on the distribution information of substation equipment corresponding to each health status value exceeding the health status evaluation index data.

[0037] Furthermore, in the intelligent monitoring and health assessment method for the entire life cycle of substation equipment, relevant maintenance plans are generated based on early warning information combined with aging evolution factor data in the target area, specifically including: The system statistically analyzes the aging evolution factor data of substation equipment that generates early warning information in the target area to determine the probability of damage for each piece of equipment. It then obtains the cost data for each damaged piece of equipment and calculates the total damage cost based on the cost data and probability of damage for each piece of equipment. Initialize the replacement plan for damaged equipment based on the total damage cost data, use the replacement plan for damaged equipment as a variable, introduce a genetic algorithm, statistically analyze the total maintenance cost data of the variable, and input the variable into the genetic algorithm for inheritance. Determine whether the total cost of the maintenance plan for the variable is greater than the preset total cost threshold for the maintenance plan. If the total cost of the maintenance plan for the variable is not greater than the preset total cost threshold for the maintenance plan, output the current variable. If the total cost of a variable's maintenance plan exceeds the preset threshold for the total cost of the maintenance plan, the variable will continue to be inherited until it does not exceed the preset threshold for the total cost of the maintenance plan.

[0038] It should be noted that the product of the cost data of the damaged equipment and the damage probability value is the total damage cost data. This method can ensure that the total cost data of the maintenance plan is not greater than the preset total cost threshold data of the maintenance plan, thereby reducing maintenance costs.

[0039] In addition, this method also includes: A new layer is constructed in the federated learning model, and the feature center of each data category is learned in the new layer. When the substation equipment data is trained locally, the model parameters are updated, and the corresponding type of the local data is also updated. The local data is updated using momentum updates, clustering, or robust aggregation on the server, and the updated global data type is then retransmitted to the substation equipment. Calculate the Mahalanobis distance between the local data of the substation equipment and the updated global data type, and encourage the local data of the substation equipment to move closer to the global data type whose Mahalanobis distance is lower than the Mahalanobis distance threshold; Local data located far from substation equipment is close to global data types whose Mahalanobis distance is below the Mahalanobis distance threshold.

[0040] It should be noted that by learning the feature center of each data category in the new layer, the category centers in the feature space are aligned, and the Non-IID problem is solved at the feature level. This decouples the uneven label distribution and feature distribution drift, which is more effective than simply aligning at the output layer. Asynchronous updates reduce the dependence on the synchronization of substation equipment prototypes.

[0041] like Figure 2 As shown, the second aspect of the present invention provides a substation equipment full life cycle intelligent monitoring and health assessment system, including a memory and a processor. The memory includes a substation equipment full life cycle intelligent monitoring and health assessment method program. When the substation equipment full life cycle intelligent monitoring and health assessment method program is executed by the processor, it implements any of the steps of the substation equipment full life cycle intelligent monitoring and health assessment method.

[0042] like Figure 3 As shown, the third aspect of the present invention provides an intelligent monitoring and health assessment device for the entire life cycle of substation equipment, comprising: Data processing module 10 is used to set up data acquisition equipment at the equipment end, collect aging evolution factor data and performance characteristic data of substation equipment through the data acquisition equipment, and process the data based on federated learning. The model building module 20 is used to collect aging evolution factor data and substation equipment performance characteristic data after being processed by federated learning, and to build a life cycle monitoring model for substation equipment. The early warning module 30 is used to obtain the health information of each substation device through the life cycle monitoring model of substation equipment, and to issue early warnings based on the health information of each substation device, generating early warning information. The maintenance plan generation module 40 is used to generate relevant maintenance plans based on early warning information and aging evolution factor data in the target area.

[0043] The fourth aspect of the present invention provides a computer-readable storage medium, including a method program for intelligent monitoring and health assessment of the entire life cycle of substation equipment. When the method program is executed by a processor, it implements any one of the steps of the method for intelligent monitoring and health assessment of the entire life cycle of substation equipment.

[0044] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0045] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0046] In addition, in the various embodiments of the present invention, each functional unit can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0047] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0048] Alternatively, if the integrated units of this invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks.

[0049] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for intelligent monitoring and health assessment of substation equipment throughout its entire lifecycle, characterized in that, Includes the following steps: Data acquisition devices are set up at the equipment end to collect aging evolution factor data and performance characteristic data of substation equipment, and the data is processed based on federated learning. Collect aging evolution factor data and substation equipment performance characteristic data after processing by federated learning, and construct a life cycle monitoring model for substation equipment. The health information of each substation device is obtained through the life cycle monitoring model of the substation equipment, and early warning information is generated based on the health information of each substation device. Based on the early warning information and the aging evolution factor data in the target area, a relevant maintenance plan is generated.

2. The intelligent monitoring and health assessment method for the entire life cycle of substation equipment according to claim 1, characterized in that, Data acquisition equipment is installed at the equipment end to collect aging evolution factor data and performance characteristic data of substation equipment, specifically including: Configure data acquisition nodes on the local end, collect data samples from the data acquisition nodes, obtain the computing power data of each data acquisition node, and initialize the size information of the collected data according to the computing power data of the data acquisition nodes. Data is collected through data acquisition nodes based on the size information of the collected data. The collected data is then subjected to knowledge distillation and input into the classroom network to obtain knowledge feature representations. Obtain the data size information of the knowledge feature representation, and determine whether the data size information of the knowledge feature representation is greater than the computing power data of the data acquisition node; When the data size information represented by the knowledge feature is greater than the computing power data of the data acquisition node, the size information of the acquired data is reduced until it is no greater than the computing power data of the data acquisition node, and the data is stored at the local end of the substation equipment.

3. The intelligent monitoring and health assessment method for the entire life cycle of substation equipment according to claim 1, characterized in that, Data processing based on federated learning specifically includes: A federated learning model is constructed to obtain the distribution information and data update frequency of local data of substation equipment, and importance evaluation data features are set according to the distribution information and data update frequency of the local data of substation equipment. The federated learning model is used to learn the data of each substation. The quantization bit width of each layer or parameter is dynamically adjusted according to the statistical characteristics of the model update. A lightweight autoencoder is used to update and compress the high-dimensional data into a low-dimensional vector. After aggregating the low-dimensional representation, it is decoded back to the model for updating and deployed on the server as part of the global model. The local computational load and communication frequency of each device are dynamically adjusted according to the status of the substation equipment, the importance of the data, and the current model status. The potential contribution of the local data distribution of the equipment to the improvement of the current global model is evaluated, and substation equipment with a contribution value higher than the contribution value threshold is scheduled for local training. Different local iteration numbers are allocated according to the status of the substation equipment.

4. The method for intelligent monitoring and health assessment of substation equipment throughout its entire lifecycle according to claim 1, characterized in that, Data on aging evolution factors processed through federated learning and performance characteristic data of substation equipment were collected, and a life cycle monitoring model for substation equipment was constructed, specifically as follows: The data on aging evolution factors processed by federated learning and the performance characteristics data of substation equipment are collected. The substation equipment is treated as a node, the electrical connection as an edge, and the aging evolution factor data as a graph attribute. Historical fault data is collected based on the performance characteristic data of the substation equipment, the fault types of the historical fault data are obtained, and the impact weights on the lifespan of the substation equipment are dynamically adjusted based on the fault types. The nodes, edges, and graph attributes are connected to construct a topology graph. A lifecycle monitoring model for substation equipment is built based on a deep learning network. The topology graph is then input into the lifecycle monitoring model for substation equipment for training. The lifespan status of substation equipment is used as the output of the model, and the model parameters are continuously updated. When the prediction accuracy of the model reaches the preset prediction accuracy, the model parameters of the lifespan monitoring model of substation equipment are saved.

5. The method for intelligent monitoring and health assessment of substation equipment throughout its entire lifecycle according to claim 1, characterized in that, The health information of each piece of substation equipment is obtained through the life cycle monitoring model of the substation equipment, specifically as follows: The aging evolution factor data and performance characteristic data of the substation equipment within a preset time are obtained and input into the life cycle monitoring model of the substation equipment for prediction, so as to obtain the life cycle status of each substation equipment. The health information of each substation device is determined based on its lifespan status, and the health information of each substation device is output.

6. The method for intelligent monitoring and health assessment of substation equipment throughout its entire lifecycle according to claim 1, characterized in that, Based on the health information of each substation device, an early warning is generated, specifically including: Set health evaluation index data, and determine whether the health information of each substation device is greater than the health evaluation index data; When the health information of the substation equipment is greater than the health evaluation index data, the distribution information of the substation equipment corresponding to each health information greater than the health evaluation index data is obtained. Early warning information is generated based on the distribution information of substation equipment corresponding to each health status information that is greater than the health status evaluation index data.

7. The method for intelligent monitoring and health assessment of substation equipment throughout its entire lifecycle according to claim 1, characterized in that, Based on the early warning information and the aging evolution factor data in the target area, a relevant maintenance plan is generated, specifically including: The damage probability value of each substation equipment that generates early warning information is obtained under the aging evolution factor data of the target area. The cost data information of each damaged equipment is obtained, and the total damage cost data is calculated based on the cost data information of each damaged equipment and the damage probability value. The replacement plan for the damaged equipment is initialized based on the total damage cost data. The replacement plan for the damaged equipment is used as a variable and introduced into a genetic algorithm. The total maintenance cost data of the variable is calculated and then input into the genetic algorithm for inheritance. Determine whether the total maintenance cost data of the variable is greater than the preset total maintenance cost threshold data. If the total maintenance cost data of the variable is not greater than the preset total maintenance cost threshold data, output the current variable. If the total cost of the maintenance plan for the variable is greater than the preset total cost threshold for the maintenance plan, the variable will continue to be inherited until it is no greater than the preset total cost threshold for the maintenance plan.

8. A smart monitoring and health assessment system for the entire lifecycle of substation equipment, characterized in that, The device includes a memory and a processor. The memory includes a program for intelligent monitoring and health assessment of substation equipment throughout its entire life cycle. When the processor executes the program, it implements the steps of the intelligent monitoring and health assessment method for substation equipment throughout its entire life cycle as described in any one of claims 1-7.

9. Intelligent monitoring and health assessment equipment for the entire life cycle of substation equipment, characterized in that, include: The data processing module is used to set up data acquisition devices on the equipment side, collect aging evolution factor data and performance characteristic data of substation equipment through the data acquisition devices, and process the data based on federated learning. The model building module is used to collect aging evolution factor data and substation equipment performance characteristic data after being processed by federated learning, and to build a life cycle monitoring model for substation equipment. The early warning module is used to obtain the health information of each substation device through the life cycle monitoring model of the substation equipment, and to issue early warnings based on the health information of each substation device, thereby generating early warning information. The maintenance plan generation module is used to generate relevant maintenance plans based on the early warning information and the aging evolution factor data in the target area.

10. A computer-readable storage medium, characterized in that, The method includes a program for intelligent monitoring and health assessment of substation equipment throughout its entire lifecycle. When the program is executed by a processor, it implements the steps of the intelligent monitoring and health assessment method for the entire lifecycle of substation equipment as described in any one of claims 1-7.

Images