Intelligent high-voltage switch cabinet monitoring system and method based on Internet of Things

Abstract

The invention relates to the technical field of switch cabinet monitoring, in particular to an intelligent high-voltage switch cabinet monitoring system and method based on the Internet of Things, and the system comprises a distributed model training module which is used for independently training a prediction model by each edge node through employing locally collected monitoring data, and uploading a training result to a cloud end in an encryption manner; the cloud generates a global model by aggregating the training results uploaded by the edge nodes; the dynamic resource negotiation module is used for dynamically adjusting a distribution strategy of monitoring resources according to the real-time load state and the fault risk level of each device; and the contribution degree quantification module is used for calculating the contribution weight and dynamically distributing the collaboration income according to the resource quantity input by each edge node in the collaboration process and the fault suppression effect. According to the invention, individual combat of a plurality of high-voltage switch cabinets is converted into intelligent cooperation, and the situation that a single equipment fault diffuses into regional power grid paralysis is effectively prevented. The power grid safety requirement is met; resource distribution is adjusted in real time along with the power grid state, and waste is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of switchgear monitoring, and specifically to an intelligent high-voltage switchgear monitoring system and method based on the Internet of Things. Background Art

[0002] High-voltage switchgears play an important protective role in the power grid. When a fault occurs in the equipment or line in the power grid, the high-voltage switchgear can quickly take countermeasures to prevent the fault from expanding, thereby protecting the normal operation of other equipment in the power grid.

[0003] In the power grid, high-voltage switchgears are usually deployed in clusters. Due to the requirement of data confidentiality in the operation of the power grid, existing monitoring systems mostly operate independently. Each switchgear corresponds to its own monitoring system separately, and the data is not shared, which may lead to a chain reaction (such as overload transfer, voltage dip) caused by the failure of a single device, and lack the ability of collaborative early warning and dynamic resource allocation among devices. Summary of the Invention

[0004] The purpose of the present invention is to provide an intelligent high-voltage switchgear monitoring system and method based on the Internet of Things to solve the problems raised in the prior art.

[0005] To achieve the above purpose, the present invention provides the following technical solution: An intelligent high-voltage switchgear monitoring system based on the Internet of Things, the monitoring system includes: Distributed model training module: Edge nodes are deployed on multiple high-voltage switchgears. Each edge node independently trains a prediction model using locally collected monitoring data and uploads the training result to the cloud in an encrypted form; the cloud generates a global model by aggregating the training results uploaded by each edge node and distributes it to each edge node to update the local model; Dynamic resource negotiation module: Communicatively connected to the distributed model training module, used to establish cooperation rules among multiple devices, and dynamically adjust the allocation strategy of monitoring resources according to the real-time load status and fault risk level of each device. The monitoring resources include communication bandwidth, computing resources, and task priorities; Low-latency collaborative network module: Connects each high-voltage switchgear using a time-sensitive communication protocol, supporting the synchronous transmission of device status data and cooperation instructions; Contribution quantification module: Calculates the contribution weight according to the amount of resources invested by each edge node during the cooperation process and the fault suppression effect, and dynamically allocates cooperation benefits.

[0006] In the distributed model training module: The training result uploaded by each edge node is the updated amount of model parameters processed by homomorphic encryption, ensuring that the original monitoring data does not leave the local; The cloud eliminates the impact of data distribution differences between different devices on the global model through the adversarial feature alignment method; The complex model generated by the cloud is converted into a lightweight version adapted to the computing power of edge nodes through model compression technology.

[0007] In the dynamic resource negotiation module: The generation of the resource allocation strategy includes the following steps: Construct a multi-dimensional state space including device load rate, health status of adjacent devices, and communication quality; Define a set of executable actions, including increasing the local data collection frequency, requesting external computing support, or releasing idle resources; Establish a reward mechanism based on historical cooperation experience, and preferentially reward strategies that can improve local monitoring accuracy and reduce global failure risks; Through iterative optimization, each device autonomously achieves a balanced state of resource allocation.

[0008] The low-latency collaborative network further includes: A device credibility evaluation mechanism based on historical communication records, which automatically downgrades nodes with frequent transmission failures; Adopt a chained data verification method to ensure the integrity and source authenticity of cooperation instructions.

[0009] In the contribution quantification module: The calculation logic of the contribution weight includes: Statistically calculate the computing resources, communication bandwidth, and key early warning data volume provided by each device during the cooperation process; Evaluate the improvement amplitude of the device's ability to suppress faults before and after participating in cooperation; Dynamically correct the basic contribution value according to the resource cost weight and the severity level of the fault; Map the corrected contribution value to the resource income distribution weight in proportion.

[0010] An intelligent high-voltage switch cabinet monitoring method based on the Internet of Things, the monitoring method includes the following steps: Step S1, deploy an edge node for each high-voltage switch cabinet, and each edge node uses local data to train and generate a local monitoring model, and encrypt and upload the local monitoring model to the cloud to generate a global model; Step S2, when any edge node detects an abnormal situation, initiate a multi-device resource negotiation request, and select the optimal resource allocation plan according to the real-time status data of each edge node; Step S3, quantify the resource input and fault suppression contributions of each edge node during the cooperation process, calculate the contribution weight of each edge node, and dynamically adjust the subsequent cooperation rights and interests distribution; Step S4: Send the optimal resource allocation plan to the edge node with abnormal conditions and execute it.

[0011] The step S1 includes: The edge node is a local intelligent computing device deployed on the high-voltage switchgear, which is used to execute local detection data processing, local monitoring model training and collaborative decision-making tasks, and interact with the cloud system. The local intelligent computing device includes a data acquisition interface, an embedded processor and a communication unit. The input of the local monitoring model is the local detection data of the high-voltage switchgear, and its output is the fault type of the high-voltage switchgear. The local detection data includes electrical parameters, temperature parameters, mechanical parameters, insulation parameters and environmental parameters. The fault types include overload, insulation aging and partial discharge. The global model is used to integrate the characteristic laws of the local monitoring data of each edge node to form unified monitoring knowledge, and update the global model periodically.

[0012] The step S2 includes: Step S201: When the local detection data of an edge node triggers a threshold, it is marked as an abnormal condition and a multi-device resource negotiation request is sent. Step S202: After the cloud receives a multi-device resource negotiation request sent by an edge node, it obtains the fault type of the abnormal condition and generates a negotiation request message. The negotiation request message includes: Required resource types: computing resources, communication bandwidth, sensor sampling frequency; Urgency level: Graded according to the fault type; Expected collaboration duration: The fault development time window estimated according to the prediction model; Step S203: Send requests to adjacent edge nodes and the cloud through a low-latency collaborative network. After receiving the requests, each edge node calculates the idle resources of its own device. The idle resources include the available rate of computing resources, the remaining communication bandwidth, and the sensor load. Generate a response message according to the idle resources. The response message includes: available required resource types, quantities and its own status labels. The own status labels include high-reliability nodes and low-load nodes. Step S204: Collect all response messages, calculate the weighted scores of each edge node that feeds back the response messages. The calculation method of the weighted score is: (computing resources × computing power weight + bandwidth × bandwidth weight) × reliability coefficient / response delay; where the response delay is the duration from when each edge node receives the request until it feeds back the response message. Arrange the weighted scores in descending order to form a node sequence, and send common collaboration instructions to each edge node in turn from top to bottom according to the node sequence. The edge node that first confirms the response to the common collaboration instruction is used as the collaboration node, and the collaboration node is specified to run a specific analysis task.

[0013] The calculation process of calculating the contribution weight of edge nodes in step S3 is as follows: Step S301: Extract the records of the common collaboration instructions completed by the edge nodes as collaboration nodes, and obtain the contribution data made by the edge nodes in the common collaboration instructions. The contribution data includes the amount of computing resources invested, the bandwidth occupancy value, and the response timeliness. The response timeliness is the delay time from receiving the request to executing the common collaboration instruction; Perform normalization processing on the contribution data, normalize the data of the amount of computing resources invested, the bandwidth occupancy value, and the response timeliness into scores in the range of 0 to 1 respectively. The computing resource and bandwidth ratio are directly mapped according to the actual value, and the response delay score increases as the delay shortens; Perform weighted summation on the scores of the amount of computing resources invested, the bandwidth occupancy value, and the response timeliness to generate a resource investment score, where the weight of computing resources is the highest, the communication bandwidth is the second, and the response delay weight is the lowest; Step S302: Select evaluation indicators according to the type of fault: Overload - type faults: Use the load reduction ratio and the number of affected devices as indicators; Insulation - type faults: Use the discharge intensity reduction ratio and the early warning time in advance as indicators; Normalize and then perform weighted summation on the performance of each edge node under different indicators to generate a fault suppression contribution value. In overload - type faults, the load reduction weight is higher than the influence range, and in insulation - type faults, the discharge suppression weight is higher than the early warning timeliness; Step S303: According to the current stability level of the power grid, set the fusion ratio of resource investment and fault suppression: The lower the stability, the higher the proportion of the fault suppression contribution; Fuse the resource investment score and the fault suppression contribution value according to the ratio to generate a comprehensive contribution degree; Convert the comprehensive contribution degree of each node into a percentage of the total contribution degree as the rights and interests distribution weight for the subsequent collaboration cycle.

[0014] In step S4, the process of executing the optimal resource allocation scheme is as follows: Select the corresponding level channel to send control instructions according to the urgency of the fault type; After receiving the instructions, the collaboration nodes execute specific analysis tasks and monitor the execution status of the instructions in real - time.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. The present invention transforms multiple high - voltage switchgears from "single - soldier operation" to "intelligent collaboration", effectively preventing the spread of single - device faults to regional power grid paralysis, and having a privacy protection function, so that data does not leave the local area, meeting the power grid security requirements; The resource allocation is adjusted in real - time according to the power grid state, avoiding waste; It has self - learning ability, and the equipment becomes more intelligent with use, without the need for manual frequent adjustment of strategies. Brief Description of the Drawings

[0016] Figure 1 This is a schematic flowchart of a method for monitoring intelligent high-voltage switchgear based on the Internet of Things according to the present invention. Detailed implementation manners

[0017] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0018] Embodiment: As Figure 1 shown, the present invention provides an intelligent high-voltage switchgear monitoring system based on the Internet of Things. The monitoring system includes: Distributed model training module: Edge nodes are deployed on multiple high-voltage switchgears. Each edge node independently trains a prediction model using locally collected monitoring data and uploads the training results to the cloud in an encrypted form; the cloud generates a global model by aggregating the training results uploaded by each edge node and distributes it to each edge node to update the local model; Dynamic resource negotiation module: Communicatively connected to the distributed model training module, used to establish cooperation rules among multiple devices, and dynamically adjust the distribution strategy of monitoring resources according to the real-time load status and fault risk level of each device. The monitoring resources include communication bandwidth, computing resources, and task priorities; Low-latency collaborative network module: Connects each high-voltage switchgear using a time-sensitive communication protocol, supporting the synchronous transmission of device status data and cooperation instructions; Contribution quantification module: Calculates the contribution weight and dynamically distributes cooperation benefits according to the amount of resources invested by each edge node during the cooperation process and the fault suppression effect.

[0019] In the distributed model training module: The training results uploaded by each edge node are the updated amounts of model parameters processed by homomorphic encryption, ensuring that the original monitoring data does not leave the local; The cloud eliminates the influence of data distribution differences between different devices on the global model through an adversarial feature alignment method; The complex model generated by the cloud is converted into a lightweight version adapted to the computing power of the edge node through model compression technology.

[0020] In the dynamic resource negotiation module: The generation of the resource allocation strategy includes the following steps: Construct a multi-dimensional state space including device load rate, health status of adjacent devices, and communication quality; Define a set of executable actions, including increasing the local data collection frequency, requesting external computing support, or releasing idle resources; Establish a reward mechanism based on historical collaboration experience, and preferentially reward strategies that can improve local monitoring accuracy and reduce global failure risks; Through iterative optimization, each device autonomously achieves a balanced state of resource allocation.

[0021] The low-latency collaborative network further includes: A device credibility evaluation mechanism based on historical communication records that automatically downgrades the weight of nodes with frequent transmission failures; Adopt a chained data verification method to ensure the integrity and source authenticity of collaborative instructions.

[0022] In the contribution quantification module: The calculation logic of the contribution weight includes: Statistically calculate the computing resources, communication bandwidth, and key early warning data volume provided by each device during the collaboration process; Evaluate the improvement amplitude of the failure suppression effect before and after the device participates in the collaboration; Dynamically correct the basic contribution value according to the resource cost weight and the severity level of the failure; Map the corrected contribution value to the resource revenue distribution weight in proportion.

[0023] An Internet of Things-based intelligent high-voltage switchgear monitoring method, the monitoring method includes the following steps: Step S1: Deploy edge nodes for each high-voltage switchgear. Each edge node uses local data to train and generate a local monitoring model, and encrypts and uploads the local monitoring model to the cloud to generate a global model; Step S2: When any edge node detects an abnormal situation, initiate a multi-device resource negotiation request, and select the optimal resource allocation plan according to the real-time status data of each edge node; Step S3: Quantify the resource input and failure suppression contributions of each edge node during the collaboration process, calculate the contribution weight of each edge node, and dynamically adjust the subsequent collaboration benefit distribution; Step S4: Send the optimal resource allocation plan to the edge node where the abnormal situation occurs and execute it.

[0024] The step S1 includes: The edge node is a local intelligent computing device deployed on the high-voltage switchgear, which is used to execute local detection data processing, local monitoring model training, and collaborative decision-making tasks, and interact with the cloud system. The local intelligent computing device includes a data acquisition interface, an embedded processor, and a communication unit; The input of the local monitoring model is the local detection data of the high-voltage switchgear, and its output is the fault type of the high-voltage switchgear; the local detection data includes electrical parameters, temperature parameters, mechanical parameters, insulation parameters, and environmental parameters; the fault types include overload, insulation aging, and partial discharge; The global model is used to integrate the characteristic laws of the local monitoring data of each edge node to form unified monitoring knowledge, and update the global model periodically.

[0025] The step S2 includes: Step S201: When the local detection data of a certain edge node triggers the threshold, it is marked as an abnormal situation, and a multi-device resource negotiation request is sent; Step S202: After the cloud receives the multi-device resource negotiation request sent by a certain edge node, it obtains the fault type of the abnormal situation and generates a negotiation request message, and the negotiation request message includes: Required resource type: computing resource, communication bandwidth, sensor sampling frequency; Urgency level: classified according to the fault type (e.g., overload = emergency level, partial discharge = high-risk level); Expected cooperation duration: the time window for fault development estimated according to the prediction model; Step S203: Send requests to adjacent edge nodes and the cloud through the low-latency cooperation network. After receiving the requests, each edge node calculates the idle resources of its own device, and the idle resources include the available rate of computing resources, the remaining communication bandwidth, and the sensor load; generate a response message according to the idle resources, and the response message includes: the available required resource type, quantity, and its own status label, and the own status label includes high-reliability nodes and low-load nodes; Step S204: Collect all response messages, calculate the weighted scores of each edge node that feeds back the response messages, and the calculation method of the weighted score is: (computing resource × computing power weight + bandwidth × bandwidth weight) × reliability coefficient / response delay; where the response delay is the duration from when each edge node receives the request until it feeds back the response message; arrange the weighted scores in descending order to form a node sequence, and send common cooperation instructions to each edge node in turn from top to bottom according to the node sequence, and take the edge node that first confirms the response to the common cooperation instruction as the cooperation node, and specify the cooperation node to run a specific analysis task.

[0026] The calculation process of calculating the contribution weight of the edge node in the step S3 is: Step S301: Extract the edge nodes as the records of the common collaboration instructions completed by the collaboration nodes, and obtain the contribution data made by the edge nodes in the common collaboration instructions. The contribution data includes the amount of computing resources invested, the bandwidth occupancy value, and the response timeliness. The response timeliness is the delay time from receiving the request to executing the common collaboration instruction; Perform normalization processing on the contribution data, normalize the data of the amount of computing resources, the bandwidth occupancy value, and the response timeliness into scores in the range of 0 to 1 respectively. The ratio of computing resources to bandwidth occupancy is directly mapped according to the actual value, and the response delay score increases as the delay shortens; Perform weighted summation on the scores of the amount of computing resources, the bandwidth occupancy value, and the response timeliness to generate a resource investment score, where the weight of computing resources is the highest, the communication bandwidth is the second, and the response delay weight is the lowest; Step S302: Select evaluation indicators according to the type of fault: Overload - type faults: Use the load reduction ratio and the number of affected devices as indicators; Insulation - type faults: Use the discharge intensity reduction ratio and the early warning advance duration as indicators; Normalize and then perform weighted summation on the performance of each edge node under different indicators to generate a fault suppression contribution value. In overload - type faults, the weight of load reduction is higher than the influence range, and in insulation - type faults, the weight of discharge suppression is higher than the early warning timeliness; Step S303: Set the fusion ratio of resource investment and fault suppression according to the current stability level of the power grid: The lower the stability, the higher the proportion of the fault suppression contribution; Fuse the resource investment score and the fault suppression contribution value according to the ratio to generate a comprehensive contribution degree; Convert the comprehensive contribution degree of each node into a percentage of the total contribution degree as the weight for benefit distribution in the subsequent collaboration cycle.

[0027] For example: The temperature of switchgear A soars to 90°C (threshold 80°C) due to excessive load.

[0028] A requests assistance from neighbors B and C: "I need more bandwidth to upload data. Who is willing to help me?" B responds: "I can provide bandwidth, but I need you to give priority to analyzing vibration data for me next time." C does not respond (its own load has reached 90%).

[0029] Collaborative execution: B allocates the idle bandwidth to A, and A quickly uploads data to the cloud for analysis.

[0030] The cloud model determines that it is necessary to reduce the load of A and automatically switches part of the current to other lines.

[0031] Contribution distribution: Among the benefits of avoiding power outages, A (detecting problems) accounts for 40%, and B (providing resources) accounts for 60%. In step S4, the process of executing the optimal resource allocation plan is as follows: Select the corresponding level channel to send control instructions according to the urgency of the fault type; After receiving the instructions, the cooperation nodes perform specific analysis tasks and monitor the execution status of the instructions in real time.

[0032] Example: Each switchgear uses its own data (such as temperature, current) to train an AI small model (such as a model for predicting faults); Encrypted upload: Only encrypt the "learning results" (gradient parameters) of the model and upload them to the cloud, without transmitting the original data; Global integration: The cloud aggregates the learning results of all models, generates a more powerful global model, and then distributes it to all switchgears; When switchgear A detects overheating, it requires more bandwidth to upload data.

[0033] At this time, switchgear B has a low load and can sell its idle bandwidth to switchgear A in exchange for priority assistance from switchgear A when switchgear B needs it in the future; Record contributions: In each cooperation, record the resources invested by each switchgear (such as how much computing power is provided).

[0034] Calculate weights: Use the Shapley value algorithm (a classic method in cooperative game theory) to allocate benefits according to historical contributions.

[0035] Example: Switchgears A, B, and C cooperated to avoid a power outage: A provided key data (contributing 50%), B provided computing power (30%), and C provided bandwidth (20%).

[0036] The benefits (such as cost savings in maintenance) are distributed in the ratio of 5:3:2.

[0037] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be included in the present invention. Any reference signs in the claims should not be regarded as limiting the claims involved.

Claims

1. An intelligent high-voltage switchgear monitoring system based on the Internet of Things, characterized in that: The monitoring system includes: Distributed model training module: Edge nodes are deployed on multiple high-voltage switchgears. Each edge node independently trains a prediction model using locally collected monitoring data and uploads the training results to the cloud in an encrypted form. The cloud generates a global model by aggregating the training results uploaded by each edge node and distributes it to each edge node to update the local model. Dynamic resource negotiation module: Communicatively connected to the distributed model training module, it is used to establish cooperation rules among multiple devices and dynamically adjust the allocation strategy of monitoring resources according to the real-time load status and fault risk level of each device. The monitoring resources include communication bandwidth, computing resources, and task priorities. Low-latency collaborative network module: Connects each high-voltage switchgear using a time-sensitive communication protocol, supporting the synchronous transmission of device status data and cooperation instructions. Contribution quantification module: Calculates the contribution weight and dynamically allocates cooperation benefits according to the amount of resources invested by each edge node during the cooperation process and the fault suppression effect.

2. The intelligent high-voltage switch cabinet monitoring system based on the Internet of Things according to claim 1, wherein: In the distributed model training module: The training results uploaded by each edge node are the updated amounts of model parameters processed by homomorphic encryption, ensuring that the original monitoring data does not leave the local area. The cloud eliminates the impact of data distribution differences between different devices on the global model through an adversarial feature alignment method. The complex model generated by the cloud is converted into a lightweight version adapted to the computing power of edge nodes through model compression technology.

3. The intelligent high-voltage switch cabinet monitoring system based on the Internet of Things according to claim 1, wherein: In the dynamic resource negotiation module: The generation of the resource allocation strategy includes the following steps: Construct a multi-dimensional state space including device load rate, health status of adjacent devices, and communication quality. Define a set of executable actions, including increasing the local data collection frequency, requesting external computing support, or releasing idle resources. Establish a reward mechanism based on historical cooperation experience, and preferentially reward strategies that can improve local monitoring accuracy and reduce global fault risks. Through iterative optimization, each device autonomously reaches a balanced state of resource allocation.

4. The intelligent high-voltage switch cabinet monitoring system based on the Internet of Things according to claim 1, characterized in that: The low-latency collaborative network further includes: A device credibility evaluation mechanism based on historical communication records that automatically downgrades the weight of nodes with frequent transmission failures. Adopts a chained data verification method to ensure the integrity and source authenticity of cooperation instructions.

5. The intelligent high-voltage switch cabinet monitoring system based on the Internet of Things according to claim 1, wherein: In the contribution quantification module: The calculation logic of the contribution weight includes: Statistical amounts of computing resources, communication bandwidth, and key early warning data provided by each device during the cooperation process. Evaluate the improvement amplitude of the fault suppression effect before and after the device participates in the cooperation. Dynamically correct the basic contribution value according to the resource cost weight and the severity level of the fault. Map the corrected contribution value to the resource income distribution weight in proportion.

6. A monitoring method for an intelligent high-voltage switchgear based on the Internet of Things, which is applied to the intelligent high-voltage switchgear monitoring system according to any one of claims 1-5, and is characterized in that: The monitoring method includes the following steps: Step S1: Deploy edge nodes on each high-voltage switchgear. Each edge node uses local data to train and generate a local monitoring model, and encrypts and uploads the local monitoring model to the cloud to generate a global model. Step S2: When any edge node detects an abnormal situation, initiate a multi-device resource negotiation request, and select the optimal resource allocation plan according to the real-time status data of each edge node. Step S3: Quantify the resource input and fault suppression contributions of each edge node during the collaboration process, calculate the contribution weights of each edge node, and dynamically adjust the subsequent collaboration rights and interests allocation; Step S4: Send the optimal resource allocation plan to the edge node with abnormal conditions and execute it.

7. The intelligent high-voltage switch cabinet monitoring method based on the Internet of Things according to claim 6, characterized in that: The step S1 includes: The edge node is a local intelligent computing device deployed on the high-voltage switchgear, which is used to execute local detection data processing, local monitoring model training, and collaborative decision-making tasks, and interact with the cloud system. The local intelligent computing device includes a data acquisition interface, an embedded processor, and a communication unit; The input of the local monitoring model is the local detection data of the high-voltage switchgear, and its output is the fault type of the high-voltage switchgear; the local detection data includes electrical parameters, temperature parameters, mechanical parameters, insulation parameters, and environmental parameters; the fault types include overload, insulation aging, and partial discharge; The global model is used to integrate the characteristic laws of the local monitoring data of each edge node to form unified monitoring knowledge, and update the global model periodically.

8. The intelligent high-voltage switch cabinet monitoring method based on the Internet of Things according to claim 6, characterized in that: The step S2 includes: Step S201: When the local detection data of a certain edge node triggers the threshold, mark it as an abnormal situation and send a multi-device resource negotiation request; Step S202: After the cloud receives the multi-device resource negotiation request sent by a certain edge node, obtain the fault type of the abnormal situation and generate a negotiation request message. The negotiation request message includes: Required resource types: computing resources, communication bandwidth, sensor sampling frequency; Urgency level: Graded according to the fault type; Expected collaboration duration: The fault development time window estimated according to the prediction model; Step S203: Send requests to adjacent edge nodes and the cloud through a low-latency collaborative network. After receiving the requests, each edge node calculates the idle resources of its own device. The idle resources include the available rate of computing resources, the remaining communication bandwidth, and the sensor load; generate a response message according to the idle resources. The response message includes: available required resource types, quantities, and its own status tags. The own status tags include high-reliability nodes and low-load nodes; Step S204: Collect all response messages, calculate the weighted scores of each edge node that feedbacks the response messages. The calculation method of the weighted score is: (computing resources × computing power weight + bandwidth × bandwidth weight) × reliability coefficient / response delay; where the response delay is the duration from when each edge node receives the request until it feedbacks the response message; arrange the weighted scores in descending order to form a node sequence, and send collaborative instruction to each edge node in turn from top to bottom according to the node sequence. The edge node that first confirms the response to the collaborative instruction is used as the collaborative node, and the collaborative node is specified to run a specific analysis task.

9. The intelligent high-voltage switch cabinet monitoring method based on the Internet of Things according to claim 6, characterized in that: The calculation process of calculating the contribution weight of the edge node in the step S3 is: Step S301: Extract the edge nodes as the records of the common collaboration instructions completed by the collaboration nodes, and obtain the contribution data made by the edge nodes in the common collaboration instructions. The contribution data includes the amount of computing resources invested, the bandwidth occupancy value, and the response timeliness. The response timeliness is the delay time from receiving the request to executing the common collaboration instructions; Perform normalization processing on the contribution data, normalize the data of the amount of computing resources, the bandwidth occupancy value, and the response timeliness into scores in the range of 0 to 1 respectively. The computing resource and bandwidth occupancy ratio are directly mapped according to the actual value, and the response delay score increases as the delay shortens; Perform weighted summation on the scores of the amount of computing resources, the bandwidth occupancy value, and the response timeliness to generate a resource investment score, where the weight of computing resources is the highest, the communication bandwidth is the second, and the response delay weight is the lowest; Step S302: Select evaluation indicators according to the fault type: Overload - type faults: Use the load reduction ratio and the number of affected devices as indicators; Insulation - type faults: Use the discharge intensity reduction ratio and the early warning advance duration as indicators; Normalize and then perform weighted summation on the performances of each edge node under different indicators to generate a fault suppression contribution value. In overload - type faults, the weight of load reduction is higher than the influence range, and in insulation - type faults, the weight of discharge suppression is higher than the early warning timeliness; Step S303: According to the current stability level of the power grid, set the fusion ratio of resource investment and fault suppression: The lower the stability, the higher the proportion of the fault suppression contribution; Fuse the resource investment score and the fault suppression contribution value according to the ratio to generate a comprehensive contribution degree; Convert the comprehensive contribution degree of each node into a percentage of the total contribution degree as the rights and interests distribution weight for the subsequent collaboration cycle.

10. The intelligent high-voltage switch cabinet monitoring method based on the Internet of Things according to claim 6, characterized in that: In the said Step S4, the process of executing the optimal resource allocation plan is as follows: Select the corresponding level channel to send control instructions according to the urgency of the fault type; After receiving the instructions, the collaboration nodes execute specific analysis tasks and monitor the execution status of the instructions in real - time.

Images