Acquisition operation and maintenance management and control method and system based on terminal interaction
By constructing an available power unit distribution matrix and using reinforcement learning for edge collaborative nodes, the overload and resource idleness problems of electric vehicle charging facilities are solved, achieving efficient resource allocation and stable operation of the charging network.
Patent Information
- Application Number
- CN202511232781.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-01
- Publication Date
- 2025-11-25
AI Technical Summary
In the traditional operation and maintenance model, electric vehicle charging facilities are widely distributed and have diverse terminal types, lacking local interaction capabilities, which leads to overload or idle resources and makes it difficult to achieve efficient and unified management.
By constructing an available power unit distribution matrix and using edge collaborative nodes for reinforcement learning, a multi-objective optimization scheduling strategy is generated to adjust charging tasks in real time, thereby achieving load balancing and reasonable resource allocation among terminals.
It effectively reduces equipment idleness and overload, improves resource utilization, ensures the stable and efficient operation of the charging network, and provides more efficient charging services.
Smart Images
Figure CN121012805A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of operation and maintenance management, more particularly, the present application relates to a terminal interaction-based operation and maintenance management method and system. BACKGROUND
[0002] Under the background of global advocacy of green travel and sustainable development, the electric vehicle industry is booming. In recent years, the sales of electric vehicles have continued to rise, and the market share has broken records. This rapid growth trend has also increased the demand for supporting facilities such as charging piles.
[0003] As the number of charging piles continues to increase in cities, parking lots, and highway service areas, the drawbacks of traditional operation and maintenance management models have gradually become apparent. The previous reliance on regular manual inspections and user feedback has not only been inefficient, but also difficult to detect potential faults and hidden dangers in real time, resulting in long-term non-use of some charging piles after failure, which seriously affects user experience. At the same time, the lack of uniformity in technical standards and communication protocols for charging piles from various manufacturers has further increased the difficulty of centralized management and unified operation and maintenance.
[0004] The above disclosed technical solutions have at least the following technical problems: electric vehicle charging facilities are widely distributed, and there are various types of terminals (such as fast charging piles, slow charging piles, and battery swap stations), traditional operation and maintenance rely on manual inspection or single cloud centralized management, and there is a lack of local interaction capability among multiple terminals (such as adjacent charging piles cannot share load status), resulting in overload or resource idling. To solve the above problems, the present application provides a solution. SUMMARY
[0005] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present application provide a terminal interaction-based operation and maintenance management method and system, which generates a multi-objective optimization scheduling strategy by constructing a usable power unit distribution matrix, to solve the problem of lack of local interaction capability among multiple terminals, resulting in overload or resource idling.
[0006] To achieve the above-mentioned purposes, the present application provides the following technical solutions: The terminal-interaction-based data collection, operation, and management method includes the following steps: acquiring the location of each charging terminal, generating a dynamic topology map, and selecting edge collaboration nodes; each terminal acquiring its own load rate, battery demand type, and grid electricity price signal in real time, abstracting the available charging capacity into virtual resource units, and reporting to the edge collaboration nodes; the edge collaboration nodes generating a multi-objective optimization scheduling strategy based on a reinforcement learning model, according to the real-time resource unit distribution, user priority, and grid constraints; splitting the charging tasks of overloaded terminals into sub-tasks, migrating them to idle terminals in the virtual resource pool according to the scheduling strategy, and synchronously updating the load status in the topology map; encrypting the scheduling records and terminal status change data and writing them into the blockchain, the cloud platform training a reinforcement learning model based on the on-chain data and distributing it to the edge nodes for iterative updates.
[0007] In a preferred embodiment, the steps of obtaining the location of each charging terminal, generating a dynamic topology map, and selecting edge collaboration nodes specifically involve: establishing wireless communication connections between the charging terminals; obtaining beacon signals containing geographic location hash values that are periodically broadcast by the charging terminals; constructing an adjacency matrix by calculating the relative distance of the receiving charging terminals based on signal strength and delay; performing eigenvalue analysis on the adjacency matrix; and selecting the charging terminal with the largest communication coverage and the lowest load rate as the edge collaboration node based on the eigenvalue analysis results; and establishing a backup mechanism to switch to the backup node when the primary edge collaboration node fails.
[0008] In a preferred embodiment, the charging terminal receiving the beacon calculates the relative distance based on signal strength and delay, and constructs an adjacency matrix. Specifically, the following steps are taken: the number of charging terminals is obtained and numbered sequentially; the distance between charging terminals is calculated based on signal strength and signal delay respectively; weight coefficients for signal strength and signal delay are obtained based on regression analysis; the relative distance between charging terminals is obtained based on weighted fusion according to the distance; adjacency matrix element rules are defined, and each element of the adjacency matrix is filled sequentially according to the relative distance.
[0009] In a preferred embodiment, each terminal obtains its own load rate and battery demand type in real time, abstracts the available charging capacity into virtual resource units, and reports them to the edge collaboration node, specifically: The system acquires the voltage and current of the charging terminal during the charging process, calculates the real-time power, and combines it with the rated power of the charging terminal to obtain the real-time load rate. It also acquires the battery's voltage, current, and remaining capacity, and by monitoring the battery voltage and current changes over time, combined with the remaining battery capacity, determines the current charging stage and demand mode of the battery. Based on the battery parameters and charging mode, the battery demand type is categorized into fast charging demand, normal charging demand, and trickle charging demand. Based on the real-time load rate and rated power, the actual charging power is calculated, and the available charging capacity is calculated based on the remaining charging time. The minimum virtual resource unit capacity is set to 1 kWh, and the available charging capacity is abstracted into several virtual resource units, which are then reported to the edge collaboration node.
[0010] In a preferred embodiment, the step of generating a multi-objective optimization scheduling strategy based on real-time resource unit distribution, user priority, and grid constraints specifically involves: acquiring power data of charging terminals within several time slices and constructing an available power unit distribution matrix; acquiring user SOC values, waiting times, and service contract level information, and calculating the user priority weight for each charging task to be scheduled; combining the user priority weights of each charging task to be scheduled to obtain a user priority weight vector for the charging task to be scheduled; acquiring grid constraint information and constructing a grid constraint set, the grid constraint information including time-of-use pricing, regional power caps, and carbon emission factors; defining a state space for reinforcement learning based on the available power unit distribution matrix, the user priority weight vector for the charging task to be scheduled, and the grid constraint set; inputting the state space into a preset deep reinforcement learning model and outputting an action space; filtering out a set of actionable actions based on the action space and action space constraints, and defining an objective reward function; and selecting the action with the largest reward value from the filtered set of actionable actions as the final scheduling strategy.
[0011] In a preferred embodiment, the step of splitting the charging task of an overloaded terminal into subtasks, migrating them to idle terminals in a virtual resource pool according to a scheduling strategy, and synchronously updating the load status in the topology map specifically involves: edge collaboration nodes monitoring the load status of each charging terminal in real time; when the load of a charging terminal exceeds a preset safety threshold, it is determined to be an overloaded terminal; obtaining the charging task currently being executed by the overloaded terminal, and obtaining the remaining power requirement, remaining charging time, and charging rate requirement of the task; splitting the charging task of the overloaded terminal into several subtasks according to the remaining power requirement, remaining charging time, and charging rate requirement of the task; edge collaboration nodes filtering out idle terminals in the virtual resource pool based on the topology map and real-time load status information; assigning the split subtasks to the filtered idle terminals based on a multi-objective optimization scheduling strategy; edge collaboration nodes sending corresponding instructions to the overloaded terminal and the idle terminals assigned subtasks to coordinate the migration of the subtasks, and the overloaded terminal stopping the execution of the corresponding subtask; after the subtask migration is completed, the edge collaboration nodes updating the load status of the overloaded terminal, adding the load status of the idle terminals of the subtasks, and updating the connection information in the topology map.
[0012] The technical effects and advantages of the data collection, operation and maintenance management method and system based on terminal interaction of this invention are as follows: 1. This invention utilizes real-time reporting of load rate, battery demand type, and grid electricity price signals from each terminal. Edge collaborative nodes construct an available power unit distribution matrix. Combining user priority weight vectors and a set of grid constraints, a multi-objective optimization scheduling strategy is generated based on a reinforcement learning model. This strategy can both rationally split and migrate overloaded terminal tasks to idle terminals and dynamically adjust based on real-time status. This effectively balances charging resource allocation, reduces equipment idleness and overload, improves the overall resource utilization of the charging network, and ensures efficient operation of the charging service.
[0013] 2. This invention utilizes terminal interaction to allow each charging terminal to collect real-time information such as its own load rate and battery demand type, and to report available charging capacity as virtual resource units. Edge collaborative nodes, based on a reinforcement learning model, generate a multi-objective optimization scheduling strategy by comprehensively considering real-time resource unit distribution, user priorities, and grid constraints. When a terminal is overloaded, tasks can be intelligently split and migrated to idle terminals, while simultaneously updating the topology map load status. This process significantly improves the rationality of charging resource allocation, reduces equipment idleness and overload, increases resource utilization, provides users with more efficient charging services, and effectively ensures the stable operation of the charging network. Attached Figure Description
[0014] Figure 1 This is a schematic diagram of the data collection, operation and maintenance management method based on terminal interaction according to the present invention.
[0015] Figure 2This is a schematic diagram of the data acquisition, operation and maintenance management system based on terminal interaction according to the present invention. Detailed Implementation
[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.
[0017] Example 1, Figure 1 The present invention provides a data collection, operation, and maintenance management method based on terminal interaction, comprising the following steps: S1, obtain the location of each charging terminal, generate a dynamic topology map, and select edge collaboration nodes; An edge collaboration node is a device or system deployed at the network edge, positioned between data sources (such as various sensors and smart terminals) and the core cloud. As a critical hub, it is responsible for collecting, processing, and analyzing data locally, and collaborating with surrounding devices to achieve more efficient, real-time business processing and decision-making, reducing reliance on the cloud and lowering data transmission latency.
[0018] The process of obtaining the location of each charging terminal, generating a dynamic topology map, and selecting edge collaboration nodes specifically involves: Establish wireless communication connections between charging terminals and obtain beacon signals containing geographic location hash values that are periodically broadcast by the charging terminals; The charging terminal receiving the beacon calculates the relative distance based on the signal strength and delay, and constructs an adjacency matrix; Eigenvalue analysis is performed on the adjacency matrix. Based on the eigenvalue analysis results, the charging terminal with the largest communication coverage and the lowest load rate is selected as the edge collaboration node. Establish a backup mechanism to switch to the backup node when the primary edge collaboration node fails.
[0019] A stronger signal and a shorter delay indicate that the two terminals are closer. These relative distances are converted into quantized data, and then an adjacency matrix is constructed. Each element in the adjacency matrix represents the connection relationship and distance information between two terminals.
[0020] The charging terminal receiving the beacon calculates the relative distance based on signal strength and delay, and constructs an adjacency matrix, specifically as follows: Obtain the number of charging terminals and number them sequentially; The distance between the charging terminals is calculated based on both signal strength and signal delay. The weighting coefficients for signal strength and signal delay are obtained based on regression analysis; The relative distance between charging terminals is obtained based on the weighted fusion of the distances mentioned above. Define the adjacency matrix element rules, and fill each element of the adjacency matrix sequentially according to the relative distance.
[0021] Adjacency matrix element rules: If no beacon signal can be received between nodes, the element is defined as 9999; for diagonal elements of the matrix, it is defined as 0.
[0022] The relative distance is specifically:
[0023]
[0024]
[0025] in, To receive signal power, For the transmitted signal power, and These represent the transmit and receive antenna gains, respectively. For the signal wavelength, The distance is measured by signal strength. It is the signal loss factor. The distance is measured by signal delay. The speed at which a signal travels through the air. For signal reception time, For signal transmission time, The distance is relative. The weighting coefficients for signal strength. This is the weighting coefficient for signal delay.
[0026] S2, each terminal obtains its own load rate, battery demand type and grid electricity price signal in real time, abstracts the available charging capacity into virtual resource units and reports them to the edge collaboration node; Each terminal obtains its own load rate and battery demand type in real time, abstracts the available charging capacity into virtual resource units, and reports them to the edge collaboration node. Specifically: The voltage and current of the charging terminal during the charging process are obtained, the real-time power is calculated, and the real-time load rate is obtained by combining the rated power of the charging terminal. The system acquires the battery's voltage, current, and remaining charge. By monitoring the changes in battery voltage and current over time and combining this with the remaining charge, it determines the battery's current charging stage and demand pattern. Based on battery parameters and charging mode, battery demand types are divided into fast charging demand, normal charging demand, and trickle charging demand. Based on the real-time load rate and rated power, the actual charging power is calculated, and the available charging capacity is calculated in combination with the current remaining charging time. The minimum virtual resource unit capacity is set to 1kWh. The available charging capacity is abstracted into several virtual resource units and reported to the edge collaboration node.
[0027] Each virtual resource unit can be considered as an independent allocable resource unit with the same value and attributes.
[0028] Furthermore, the specific steps for determining the current charging stage and demand pattern of the battery are as follows: when the battery's SOC is low, it is in the constant current charging stage, requiring a large and stable charging current; while when the battery is close to being fully charged, it enters the constant voltage charging stage, at which point the charging current gradually decreases to prevent the battery from being overcharged.
[0029] If the battery SOC is detected to be below 20% and the vehicle shows signs of needing to travel again in a short period of time (such as obtaining trip planning information through the linkage between the vehicle and the mobile APP), it is determined to be a fast charging demand type.
[0030] S3, the edge collaborative node is based on a reinforcement learning model to generate a multi-objective optimization scheduling strategy according to the real-time resource unit distribution, user priority and power grid constraints; The process of generating a multi-objective optimization scheduling strategy based on real-time resource unit distribution, user priorities, and power grid constraints is as follows: Acquire the power data of charging terminals within a certain time slice and construct the distribution matrix of available power units; Obtain the user's SOC value, waiting time, and service contract level information, and calculate the user priority weight for each charging task to be scheduled; Combine the user priority weights of each charging task to be scheduled to obtain the user priority weight vector of the charging task to be scheduled. Obtain grid constraint information and construct a set of grid constraint conditions, including time-of-use electricity pricing, regional power caps, and carbon emission factors; The state space for reinforcement learning is defined based on the distribution matrix of available power units, the user priority weight vector of the charging task to be scheduled, and the set of grid constraints. The state space is input into a pre-defined deep reinforcement learning model, and the action space is output.
[0031] Furthermore, the step of generating a multi-objective optimization scheduling strategy based on real-time resource unit distribution, user priorities, and grid constraints also includes: Based on the action space and its constraints, a set of possible actions is selected, and a target reward function is defined. From the filtered set of possible actions, select the action with the highest reward value as the final scheduling strategy; The edge collaboration node converts the action with the highest selected reward value into a specific scheduling instruction and sends it to the corresponding charging terminal; The edge collaborative node updates the available power unit distribution matrix, the status of the charging tasks to be scheduled (such as removing completed charging tasks from the scheduling list) and relevant user information (such as SOC value, waiting time, etc.) based on the information fed back by the charging terminal, thereby obtaining the new status.
[0032] The sample is stored as a training sample in the experience replay buffer. A batch of samples is periodically sampled from the experience replay buffer, and the parameters of the deep reinforcement learning model are updated using optimization algorithms such as gradient descent, enabling the model to learn better scheduling strategies.
[0033] The user priority weight for each scheduled charging task is as follows:
[0034] The target reward function is specifically as follows:
[0035] in, User priority weight, and These are the maximum and minimum values of the battery's state of charge, respectively. For user i, the battery state of charge. Let i be the waiting time. For user i, the service contract level For natural index, For the target reward function, , and These are the weighting factors for power grid stability, user satisfaction, and penalty costs, respectively. and These represent the standard deviation and mean of the non-zero elements in the available power unit distribution matrix, respectively. Number of tasks to be migrated This is the expected latency decay factor after the migration of the ii-th task. To incur penalties and costs.
[0036] The action space constraints are specifically as follows: Each task can only be migrated to one terminal:
[0037] The total power of the terminals after relocation shall not exceed the rated power:
[0038] The total power of the area shall not exceed the maximum allowable power:
[0039] in, To migrate task i to terminal j, This represents the total number of charging terminals. Number of tasks to be migrated For the power requirements of task i, The rated power of terminal j, This represents the maximum permissible power for the region.
[0040] The State of Charge (SOC) value reflects the remaining charge of a user's electric vehicle battery. The lower the SOC value, the more urgently the user may need to charge. The longer the waiting time, the higher the user's priority should be. The service contract level reflects the agreement between the user and the charging service provider. Users with higher-level contracts may have higher priority.
[0041] The action space is the set of scheduling actions that edge collaborative nodes can take. Each action represents allocating virtual resource units from one charging terminal to another, or determining the charging power adjustment of a charging terminal.
[0042] S4, split the charging task of the overloaded terminal into sub-tasks, migrate them to the idle terminal in the virtual resource pool according to the scheduling strategy, and update the load status in the topology map synchronously. The process of splitting the charging task of overloaded terminals into sub-tasks, migrating them to idle terminals in the virtual resource pool according to the scheduling strategy, and synchronously updating the load status in the topology map specifically involves: Edge collaboration nodes monitor the load of each charging terminal in real time. When the load of a charging terminal exceeds a preset safety threshold, it is determined to be an overloaded terminal. Obtain the charging task currently being executed by the overloaded terminal, and obtain the remaining power requirement, remaining charging time, and charging rate requirement of the task; The charging task of the overloaded terminal is divided into several sub-tasks based on the remaining power demand, remaining charging time, and charging rate requirements. Edge collaboration nodes filter out idle terminals in the virtual resource pool based on the topology map and real-time load status information; Based on a multi-objective optimization scheduling strategy, the split subtasks are assigned to selected idle terminals. Edge collaboration nodes send corresponding instructions to overloaded terminals and idle terminals assigned subtasks to coordinate the migration of subtasks, and the overloaded terminals stop executing the corresponding subtasks. After the subtask migration is completed, the edge collaboration node updates the load status of overloaded terminals, increases the load status of idle terminals in the subtask, and updates the connection information in the topology map.
[0043] An idle terminal should meet the following conditions: it should have enough available power units to handle the charging needs of the subtask; and it should have a good communication connection with the overloaded terminal to ensure the smooth migration and execution of the subtask. By querying the available power unit distribution matrix, the charging terminal with more available power units than the power required by the subtask within the corresponding time slice can be identified.
[0044] S5 encrypts scheduling records and terminal status change data and writes them to the blockchain. The cloud platform trains a reinforcement learning model based on the on-chain data and distributes it to edge nodes for iterative updates.
[0045] The process involves encrypting scheduling records and terminal status change data, writing them to the blockchain, and then using the cloud platform to train a reinforcement learning model based on the on-chain data and distributing it to edge nodes for iterative updates. The scheduling records (including information on the allocation, migration, and execution of charging tasks) and terminal status change data (load changes, fault information, and connection status of charging terminals) are encrypted. Edge nodes in the charging network are selected as blockchain nodes, and encrypted scheduling records and terminal status change data are used to construct blockchain transactions. The blockchain's consensus mechanism is used to verify blockchain transactions, and verified blockchain transactions are packaged into blocks and written into the blockchain. The cloud platform obtains encrypted scheduling records and terminal status change data from the blockchain; Based on scheduling records and terminal state change data, the neural network parameters of the model are updated by continuously looping through state-action-reward-next state to obtain the maximum reward function; The trained and optimized reinforcement learning model is packaged to generate a model file that runs on the edge nodes; After receiving the model file, the edge node unpacks and verifies it, and replaces the original model with the reinforcement learning model that has passed the verification.
[0046] Example 2, a data acquisition, operation and maintenance management system based on terminal interaction, includes the following modules: Terminal location acquisition module: used to acquire the location of each charging terminal, generate a dynamic topology map, and select edge collaboration nodes; Terminal resource reporting module: used by each terminal to obtain its own load rate, battery demand type and grid electricity price signal in real time, abstract the available charging capacity into virtual resource units and report to the edge collaboration node; Optimized scheduling strategy generation module: This module is used by edge collaborative nodes to generate multi-objective optimized scheduling strategies based on reinforcement learning models, real-time resource unit distribution, user priorities, and power grid constraints. Task splitting and migration module: used to split the charging tasks of overloaded terminals into subtasks, migrate them to idle terminals in the virtual resource pool according to the scheduling strategy, and synchronously update the load status in the topology map. Data encryption and model training update module: This module is used to encrypt scheduling records and terminal status change data and write them to the blockchain. The cloud platform trains reinforcement learning models based on the on-chain data and distributes them to edge nodes for iterative updates.
[0047] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.
[0048] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.
[0049] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.
[0050] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.
[0051] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.
[0052] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A data collection, operation, and maintenance management method based on terminal interaction, characterized in that, Includes the following steps: The location of each charging terminal is obtained, a dynamic topology map is generated, and edge collaboration nodes are selected; Each terminal obtains its own load rate, battery demand type and grid electricity price signal in real time, abstracts the available charging capacity into virtual resource units and reports them to the edge collaboration node; The edge collaborative nodes generate multi-objective optimization scheduling strategies based on reinforcement learning models, according to real-time resource unit distribution, user priorities, and power grid constraints. The charging task of overloaded terminals is split into subtasks, migrated to idle terminals in the virtual resource pool according to the scheduling strategy, and the load status in the topology map is updated synchronously. The scheduling records and terminal status change data are encrypted and written to the blockchain. The cloud platform trains a reinforcement learning model based on the on-chain data and distributes it to the edge nodes for iterative updates.
2. The data collection, operation, and maintenance management method based on terminal interaction according to claim 1, characterized in that, The process of obtaining the location of each charging terminal, generating a dynamic topology map, and selecting edge collaboration nodes specifically involves: Establish wireless communication connections between charging terminals and obtain beacon signals containing geographic location hash values that are periodically broadcast by the charging terminals; The charging terminal receiving the beacon calculates the relative distance based on the signal strength and delay, and constructs an adjacency matrix; Eigenvalue analysis is performed on the adjacency matrix. Based on the eigenvalue analysis results, the charging terminal with the largest communication coverage and the lowest load rate is selected as the edge collaboration node. Establish a backup mechanism to switch to the backup node when the primary edge collaboration node fails.
3. The data collection, operation, and maintenance management method based on terminal interaction according to claim 2, characterized in that, The charging terminal receiving the beacon calculates the relative distance based on signal strength and delay, and constructs an adjacency matrix, specifically as follows: Obtain the number of charging terminals and number them sequentially; The distance between the charging terminals is calculated based on both signal strength and signal delay. The weighting coefficients for signal strength and signal delay are obtained based on regression analysis; The relative distance between charging terminals is obtained based on the weighted fusion of the distances mentioned above. Define the adjacency matrix element rules, and fill each element of the adjacency matrix sequentially according to the relative distance.
4. The data collection, operation, and maintenance management method based on terminal interaction according to claim 3, characterized in that, Each terminal obtains its own load rate and battery demand type in real time, abstracts the available charging capacity into virtual resource units, and reports them to the edge collaboration node. Specifically: The voltage and current of the charging terminal during the charging process are obtained, the real-time power is calculated, and the real-time load rate is obtained by combining the rated power of the charging terminal. The system acquires the battery's voltage, current, and remaining charge. By monitoring the changes in battery voltage and current over time and combining this with the remaining charge, it determines the battery's current charging stage and demand pattern. Based on battery parameters and charging mode, battery demand types are divided into fast charging demand, normal charging demand, and trickle charging demand. Based on the real-time load rate and rated power, the actual charging power is calculated, and the available charging capacity is calculated in combination with the current remaining charging time. The minimum virtual resource unit capacity is set to 1kWh. The available charging capacity is abstracted into several virtual resource units and reported to the edge collaboration node.
5. The data collection, operation, and maintenance management method based on terminal interaction according to claim 4, characterized in that, The process of generating a multi-objective optimization scheduling strategy based on real-time resource unit distribution, user priorities, and power grid constraints is as follows: Acquire the power data of charging terminals within a certain time slice and construct the distribution matrix of available power units; Obtain the user's SOC value, waiting time, and service contract level information, and calculate the user priority weight for each charging task to be scheduled; Combine the user priority weights of each charging task to be scheduled to obtain the user priority weight vector of the charging task to be scheduled. Obtain grid constraint information and construct a set of grid constraint conditions, including time-of-use electricity pricing, regional power caps, and carbon emission factors; The state space for reinforcement learning is defined based on the distribution matrix of available power units, the user priority weight vector of the charging task to be scheduled, and the set of grid constraints. The state space is input into a pre-defined deep reinforcement learning model, and the action space is output. Based on the action space and its constraints, a set of possible actions is selected, and a target reward function is defined. From the filtered set of possible actions, the action with the highest reward value is selected as the final scheduling strategy.
6. The data collection, operation, and maintenance management method based on terminal interaction according to claim 5, characterized in that, The process of splitting the charging task of overloaded terminals into sub-tasks, migrating them to idle terminals in the virtual resource pool according to the scheduling strategy, and synchronously updating the load status in the topology map specifically involves: Edge collaboration nodes monitor the load of each charging terminal in real time. When the load of a charging terminal exceeds a preset safety threshold, it is determined to be an overloaded terminal. Obtain the charging task currently being executed by the overloaded terminal, and obtain the remaining power requirement, remaining charging time, and charging rate requirement of the task; The charging task of the overloaded terminal is divided into several sub-tasks based on the remaining power demand, remaining charging time, and charging rate requirements. Edge collaboration nodes filter out idle terminals in the virtual resource pool based on the topology map and real-time load status information; Based on a multi-objective optimization scheduling strategy, the split subtasks are assigned to selected idle terminals. Edge collaboration nodes send corresponding instructions to overloaded terminals and idle terminals assigned subtasks to coordinate the migration of subtasks, and the overloaded terminals stop executing the corresponding subtasks. After the subtask migration is completed, the edge collaboration node updates the load status of overloaded terminals, increases the load status of idle terminals in the subtask, and updates the connection information in the topology map.
7. The data collection, operation, and maintenance management method based on terminal interaction according to claim 6, characterized in that, The relative distance is specifically: in, To receive signal power, For the transmitted signal power, and These represent the transmit and receive antenna gains, respectively. For the signal wavelength, The distance is measured by signal strength. It is the signal loss factor. The distance is measured by signal delay. The speed at which a signal travels through the air. For signal reception time, For signal transmission time, The distance is relative. The weighting coefficients for signal strength. This is the weighting coefficient for signal delay.
8. The data collection, operation, and maintenance management method based on terminal interaction according to claim 7, characterized in that, The action space constraints are specifically as follows: in, To migrate task i to terminal j, This represents the total number of charging terminals. Number of tasks to be migrated For the power requirements of task i, The rated power of terminal j, This represents the maximum permissible power for the region.
9. The data collection, operation, and maintenance management method based on terminal interaction according to claim 8, characterized in that, The user priority weight for each scheduled charging task is as follows: The target reward function is specifically as follows: in, User priority weight, and These are the maximum and minimum values of the battery's state of charge, respectively. For user i, the battery state of charge. Let i be the waiting time. For user i, the service contract level For natural index, For the target reward function, , and These are the weighting factors for power grid stability, user satisfaction, and penalty costs, respectively. and These represent the standard deviation and mean of the non-zero elements in the available power unit distribution matrix, respectively. Number of tasks to be migrated This is the expected latency decay factor after the migration of the ii-th task. To incur penalties and costs.
10. A system using the terminal-interaction-based data acquisition, operation, and maintenance management method as described in any one of claims 1-9, characterized in that, Includes the following modules: Terminal location acquisition module: used to acquire the location of each charging terminal, generate a dynamic topology map, and select edge collaboration nodes; Terminal resource reporting module: used by each terminal to obtain its own load rate, battery demand type and grid electricity price signal in real time, abstract the available charging capacity into virtual resource units and report to the edge collaboration node; Optimized scheduling strategy generation module: This module is used by edge collaborative nodes to generate multi-objective optimized scheduling strategies based on reinforcement learning models, real-time resource unit distribution, user priorities, and power grid constraints. Task splitting and migration module: used to split the charging tasks of overloaded terminals into subtasks, migrate them to idle terminals in the virtual resource pool according to the scheduling strategy, and synchronously update the load status in the topology map. Data encryption and model training update module: This module is used to encrypt scheduling records and terminal status change data and write them to the blockchain. The cloud platform trains reinforcement learning models based on the on-chain data and distributes them to edge nodes for iterative updates.