Microgrid control method based on edge computing and related products

Abstract

This invention provides a microgrid control method and related products based on edge computing, relating to the field of power supply and distribution systems. The method includes: acquiring microgrid equipment operation data and control requests from an edge computing node; determining the request control task corresponding to the control request; determining the operation control task corresponding to the equipment operation data; adding the request control task and the operation control task to a task execution queue according to the pre-configured task priorities of the microgrid's control strategy; and sequentially executing the control tasks in the task execution queue by the edge computing node and / or the microgrid's cloud control platform. This invention effectively addresses the problems of control response lag and chaotic control task scheduling caused by the diverse types and complex control of microgrid equipment, improving the real-time performance, security, and coordination of microgrid control methods.

Description

Technical Field

[0001] This invention relates to the field of power supply and distribution systems, and in particular to a microgrid control method and related products based on edge computing. Background Technology

[0002] With the widespread application of renewable energy and the popularization of distributed generation systems, microgrids, as an energy application architecture that integrates power generation, energy storage, and power consumption, are gradually becoming an important part of modern power supply. Microgrids typically include various components such as photovoltaics, wind power, energy storage devices (such as batteries), loads, and possible backup generators, and use advanced control strategies to optimize energy use efficiency and improve power supply reliability and stability.

[0003] Microgrids are characterized by a wide variety of equipment types, diverse operating characteristics, and multiple control objectives, making coordinated control highly complex. Traditional microgrid control methods generally employ a combination of a unified periodic polling mechanism and an emergency alarm handling mechanism. Under normal operating conditions, the microgrid uses a periodic polling mechanism for control, which generates adjustment commands based on currently collected operating status data and a predetermined control strategy according to a set control cycle to achieve steady-state control optimization. When an emergency alarm occurs, the microgrid uses an emergency alarm handling mechanism for control, which initiates emergency control based on the alarm information and triggers protection actions.

[0004] However, the above control methods have significant limitations. On the one hand, the periodic polling mechanism and the emergency alarm handling mechanism are processed by different logical paths, lacking a unified task generation and priority scheduling framework. The system struggles to dynamically determine the execution order of the two types of control tasks, potentially leading to high-security-level tasks being blocked by low-priority optimization tasks. On the other hand, the emergency alarm handling mechanism can only handle predefined alarm information and cannot identify abnormal states hidden in the regular data stream (such as SOC mutations, accumulated communication packet loss, and device status logic conflicts), resulting in a large number of potential risks not being promptly converted into control tasks.

[0005] Furthermore, to achieve globally optimal scheduling and refined management, microgrids often need to process large amounts of data and perform complex computational tasks, such as energy management across multiple time scales and equipment health status prediction. Due to limitations in cost, physical space, and communication methods, the computing and storage resources of edge computing nodes deployed in microgrid sites are relatively limited, making it impossible to complete all computational tasks. If these tasks were entirely dependent on a cloud-based control platform, it could lead to significant latency, affecting the system's real-time response capabilities. Summary of the Invention

[0006] One object of the present invention is to provide a microgrid control method and related products based on edge computing that at least solves any of the above-mentioned technical problems.

[0007] A further objective of this invention is to solve the problem of intelligent collaborative control of microgrids in multi-device, highly complex scenarios.

[0008] Another further objective of this invention is to realize cloud-edge collaborative dynamic control of microgrids and solve the problem of computing power allocation for edge computing nodes.

[0009] Another further objective of this invention is to improve the resource computing efficiency of edge computing nodes.

[0010] Specifically, this invention provides a microgrid control method based on edge computing. The method includes: The edge computing nodes acquire equipment operation data and control requests from the microgrid. Determine the request control task corresponding to the control request; Determine the operation control tasks corresponding to the equipment operation data; Based on the pre-configured task priorities according to the microgrid's control strategy, request control tasks and operation control tasks are added to the task execution queue. The control tasks in the task execution queue are executed sequentially by the edge computing nodes and / or the cloud control platform of the microgrid.

[0011] Optionally, the steps for determining the operation control task corresponding to the equipment operation data include: Identify whether the equipment operation data includes event-triggered data that is not part of the regular data. Regular data refers to data collected from various devices in the microgrid according to a preset collection cycle. If so, the event handling task corresponding to the event trigger data will be used as the execution control task; If not, identify whether there is abnormal data in the equipment operation data. If there is abnormal data, the analysis and response task for the abnormal data will be used as the operation control task. If there is no abnormal data, the operation control task will be generated based on the regular data.

[0012] Optionally, generating runtime control tasks based on conventional data also includes: Obtain the control cycle for each pre-configured control requirement; Control tasks are generated according to the control cycle to meet control requirements and are then used as operational control tasks.

[0013] Optionally, the steps for the control tasks in the task execution queue to be executed sequentially by the edge computing nodes and / or the cloud control platform of the microgrid include: Retrieve the control task at the head of the task execution queue and use it as the target control task; The target control task is broken down into edge control sub-tasks and cloud control sub-tasks; Data processing resources of edge computing nodes are allocated according to the data processing requirements of edge control subtasks, so that edge computing nodes can perform data processing for edge control subtasks; The data transmission link is determined based on the data transmission requirements of the cloud control subtask, and the data required by the cloud control subtask is transmitted to the cloud control platform through the data transmission link.

[0014] Optionally, the step of allocating data processing resources of edge computing nodes according to the data processing requirements of edge control subtasks includes: Obtain the available computing power of edge computing nodes to determine computing power constraints; By using a mixed-integer linear programming model to optimize the computing power occupied by each functional module of the edge computing node under the condition of satisfying computing power constraints, the functional modules include: storage module, alarm processing module, control command generation module, logic judgment module, and log processing module.

[0015] Optionally, the steps for determining the data transmission link based on the data transmission requirements of the cloud control subtask include: Determine the data transmission target for the cloud control subtask; Traverse all network links to the data transmission target and obtain multiple evaluation metrics for each network link, including transmission bandwidth, transmission delay, latency jitter rate, and packet loss rate. Multiple evaluation indicators are merged using an evaluation function to obtain a comprehensive evaluation indicator; The network link with the best comprehensive evaluation index is selected as the data transmission link.

[0016] Optionally, the steps of splitting the target control task into edge control subtasks and cloud control subtasks include: Based on preset task division rules, the target control task is divided into edge control subtasks and cloud control subtasks; among which... The task division rules are based on the attributes of the subtasks, including real-time requirements, computational complexity, data dependency scope, and safety criticality level.

[0017] Optionally, the edge control subtasks include: local control, protection action processing, data preprocessing, and human-computer interaction; The cloud-based control subtasks include: deep learning model training and runtime log analysis.

[0018] Optionally, after transmitting the data required for the cloud control subtask to the cloud control platform, the following may also be included: Obtain the cloud control policy generated by the cloud control platform after executing the cloud control sub-task; The cloud-based control strategy is merged with the local control strategy obtained by the edge computing node executing edge control subtasks to generate the final control command and issue it for execution.

[0019] According to another aspect of the present invention, a computer program product is also provided, comprising a computer program that, when executed by a processor, implements the steps of any of the above-described edge computing-based microgrid control methods.

[0020] According to another aspect of the present invention, a computer-readable storage medium is also provided, on which a computer program is stored, wherein the computer program, when executed by a processor, implements the steps of any of the above-described edge computing-based microgrid control methods.

[0021] According to another aspect of the present invention, a computer device is also provided, which includes a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of any of the above-described edge computing-based microgrid control methods.

[0022] The edge computing-based microgrid control method of this invention generates request control tasks and operation control tasks respectively based on control requests and equipment operation data, and schedules their execution according to preset priorities, thus solving the problem of fragmented control in existing microgrids. By actively identifying event-triggered data and abnormal data, high-priority control tasks can be generated in a timely manner, avoiding delays in safety-critical operations. The task queue mechanism ensures that important tasks are executed first, effectively addressing the problems of control response lag and chaotic control task scheduling caused by the variety of microgrid equipment types and control complexity, thereby improving the real-time performance, security, and coordination of control.

[0023] Furthermore, the microgrid control method based on edge computing of the present invention dynamically divides the control task into two types of sub-tasks: edge and cloud. It also matches the data processing resources and transmission link selection mechanism of the edge computing node with the data processing resources of the edge computing node and the cloud respectively, making full use of the data processing advantages of the edge computing node and the cloud, realizing the complementary synergy of cloud and edge capabilities, and meeting the computing power and real-time requirements of microgrid control.

[0024] Furthermore, the microgrid control method based on edge computing of the present invention introduces a mixed integer linear programming model to dynamically optimize the computing power of edge computing nodes. Under the premise of meeting computing power constraints, it realizes precise resource allocation of modules such as storage, alarm, and control generation, avoiding resource crowding or idleness between modules caused by fixed allocation ratios, and meeting the computing power requirements under different control tasks.

[0025] Furthermore, the microgrid control method based on edge computing of the present invention constructs an evaluation function based on multi-dimensional network evaluation indicators for link optimization, so that the data transmission path selection shifts from dependence on a single indicator to multi-factor collaborative decision-making, effectively avoiding transmission interruption or strategy delay caused by link quality fluctuations, and enhancing the reliability of cloud control subtask data upload and the adaptive capability of cloud-edge communication links.

[0026] The above and other objects, advantages and features of the present invention will become more apparent to those skilled in the art from the following detailed description of specific embodiments of the invention in conjunction with the accompanying drawings. Attached Figure Description

[0027] The following sections will describe some specific embodiments of the invention in detail by way of example and not limitation, with reference to the accompanying drawings. The same reference numerals in the drawings denote the same or similar parts or portions. Those skilled in the art should understand that these drawings are not necessarily drawn to scale. In the drawings: Figure 1 This is a schematic diagram of the system architecture of a microgrid according to an embodiment of the present invention; Figure 2 This is a schematic diagram of a microgrid control method based on edge computing according to an embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the determination of control tasks in a microgrid control method based on edge computing according to an embodiment of the present invention; Figure 4 This is a schematic diagram illustrating the execution of a control task in a microgrid control method based on edge computing according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the process of allocating data processing resources for edge computing nodes in a microgrid control method based on edge computing according to an embodiment of the present invention; Figure 6 This is a schematic diagram of the functional modules of an edge computing node in a microgrid control method based on edge computing according to an embodiment of the present invention; Figure 7 This is a flowchart illustrating the process of determining a data transmission link in a microgrid control method based on edge computing according to an embodiment of the present invention. Figure 8 This is a schematic diagram of a computer program product according to an embodiment of the present invention; Figure 9 This is a schematic diagram of a computer-readable storage medium according to an embodiment of the present invention; Figure 10 This is a schematic block diagram of a computer device according to an embodiment of the present invention. Detailed Implementation

[0028] Those skilled in the art should understand that the embodiments described below are merely a part of the embodiments of the present invention, and not all of the embodiments of the present invention. These partial embodiments are intended to explain the technical principles of the present invention and are not intended to limit the scope of protection of the present invention. Based on the embodiments provided by the present invention, all other embodiments obtained by those skilled in the art without creative effort should still fall within the scope of protection of the present invention.

[0029] Figure 1 This is a schematic diagram of a microgrid system architecture according to an embodiment of the present invention. The microgrid generally includes: distributed generation equipment 11, energy storage equipment 12, public grid connection equipment 13, charging equipment 14, electrical appliances 15, edge computing nodes (or edge controllers) 21, cloud control platform 24, etc. The distributed generation equipment 11, energy storage equipment 12, public grid connection equipment 13, charging equipment 14, and electrical appliances 15 serve as terminal equipment of the microgrid.

[0030] Distributed generation equipment 11 is used to convert renewable energy sources such as solar and wind energy into electrical energy. Energy storage equipment 12 can use battery energy storage, mechanical energy storage, and thermal energy storage methods to controllably store and release electrical energy, balancing power supply and demand and smoothing output fluctuations of renewable energy. Public grid connection equipment 13 is used to connect to the public power grid, utilizing the power from the public grid to supplement the power supply capacity of distributed generation equipment 11 and energy storage equipment 12. Charging equipment 14 is a key infrastructure for charging electric vehicles and can provide charging services in a controlled manner. Electrical equipment 15 is the power supply target of the microgrid and may include heat pump systems, lighting systems, etc.

[0031] In embodiments using battery energy storage, the energy storage device 12 has a Battery Management System (BMS) for detecting the status of the energy storage battery and providing safety protection and performance optimization. The microgrid may also include a Power Conversion System (PCS) for power conversion and power regulation.

[0032] The aforementioned terminal devices are used to execute control commands issued by the edge computing node 21 and complete their respective functions. The terminal devices can provide their respective operational data to the edge computing node 21 directly or through a data acquisition device. This operational data may include: power data, switch action data, power consumption data, monitoring and alarm data, sensor data, operational status data, and log data.

[0033] Terminal equipment may also include microgrid user terminals for use by users with microgrid management authority, such as smartphones, tablets, computers, or other dedicated electronic client devices with power management application software installed. Users can use the user terminals to understand the operation of the microgrid and submit control requests. The user terminals convert the user-provided control requests into control requests and send them to the edge computing node 21.

[0034] Edge computing node 21 is set up on the local side of the microgrid to collect operating status data of devices within the control range, and supports small-scale local data lightweight processing, small amount of data storage, real-time device control, and control strategy distribution.

[0035] The cloud control platform 24 communicates with multiple edge computing nodes 21 to analyze, store, and statistically process the data uploaded by the edge computing nodes 21, supporting large-scale data analysis, deep learning model training, and large-scale data storage.

[0036] The microgrid control process can include: terminal devices uploading device operation data and control requests to edge computing nodes 21; for simple control tasks, such as instructing an air conditioner to turn off, edge computing nodes 21 directly perform calculations and issue instructions to terminal devices; for complex control tasks, such as deep learning model training, cross-regional collaborative decision-making, and big data matching, edge computing nodes 21 upload relevant data to the cloud control platform 24, and after completing the control task, the cloud control platform 24 sends the processing results to the terminal devices through edge computing nodes 21; edge computing nodes 21 are regulated by the cloud control platform 24 and correct and optimize according to the control strategies issued by the cloud control platform 24 to ensure that the control strategies are issued accurately and quickly, thereby realizing cloud-edge collaboration.

[0037] The microgrid system can also be configured with data transmission devices such as network routers, communication relay equipment, and IoT transmission equipment to provide a stable data transmission channel between devices and to realize data interaction between terminal devices, edge computing nodes 21, and cloud control platform 24 by adopting data transmission methods adapted to each device.

[0038] This embodiment also provides a microgrid control method based on edge computing, which distributes control tasks to edge computing nodes 21 and cloud control platforms 24, making full use of the respective data processing advantages of edge computing nodes 21 and cloud control platforms 24, and realizing complementary collaboration between cloud and edge.

[0039] Figure 2 This is a schematic diagram of a microgrid control method based on edge computing according to an embodiment of the present invention. The microgrid control method based on edge computing generally includes: Step S601: The edge computing node acquires the microgrid's device operation data and control requests. Device operation data may include power data, switch action data, energy consumption data, monitoring and alarm data, sensor data, operating status data, and log data for each device. Control requests may include: operating mode changes, load management, and responses to environmental changes.

[0040] Equipment operation data may include: routine data collected by edge computing nodes from various devices in the microgrid according to a preset collection cycle, and event-triggered data that is not routine data. Control requests can be issued by the controlled device terminals in the microgrid based on their own functions and operating status, or by management users through the human-machine interface of user terminals or edge computing nodes.

[0041] Step S602: Determine the request control task corresponding to the control request. Control requests generally already clearly correspond to control tasks. In the method of this embodiment, the control task corresponding to the control request is defined as a separate category, namely, a request control task. For example, if a user issues a control request for a certain electrical device, the control task is to generate control instructions for that electrical device.

[0042] Step S603: Determine the operation control task corresponding to the equipment operation data.

[0043] Step S604: Add the request control task and the operation control task to the task execution queue according to the pre-configured task priority of the microgrid control strategy.

[0044] Step S605: The edge computing nodes and / or the cloud control platform of the microgrid sequentially execute the control tasks in the task execution queue.

[0045] The task execution queue contains multiple control tasks arranged sequentially, with the order configured according to real-time requirements and importance. Tasks in the queue are executed sequentially. The process of adding request control tasks and running control tasks to the task execution queue may include: determining if there are any incomplete control tasks in the queue; if not, adding the request and running control tasks to the queue according to a preset priority rule (e.g., security control tasks > protection control tasks > optimization control tasks > periodic control tasks), and selecting the task at the head of the queue as the target control task. This ensures that important control tasks have priority in resource acquisition, and that control tasks in the queue are executed sequentially, avoiding task conflicts or delays in critical operations.

[0046] If there are still unfinished control tasks in the task execution queue, the request control task and the running control task are first compared with the existing control tasks in the task execution queue to see if the current request control task and the running control task are update tasks of the existing control tasks. If so, the update task replaces the existing control task; if the request control task and the running control task are not update tasks of the existing control tasks, the task priorities of the request control task, the running control task, and the existing control tasks are determined, and the control tasks and the running control tasks are inserted into the task execution queue according to the priorities. The aforementioned update task refers to a new version of the request control task and the running control task that belongs to the same control target as the existing task.

[0047] In other words, the task execution queue contains multiple control tasks arranged in order of priority. The task order is determined based on their real-time requirements and importance, and they are executed sequentially. If there are no unfinished control tasks in the task execution queue, the request control task and the running control task are inserted into the corresponding positions in the queue according to the preset priority rules, with the task at the head of the queue serving as the target control task (i.e., the currently pending task), ensuring that high-priority tasks receive execution resources first. If there are unfinished control tasks in the task execution queue, it is determined whether the newly generated request control task or running control task is an update task of an existing task in the queue. If it is an update task (i.e., both belong to the old and new versions of the same control target, such as updating the charging and discharging instructions for the same energy storage unit), the new task replaces the original task. If it is not an update task, the priorities of the new task and each existing task in the queue are obtained respectively, and the new task is inserted into the appropriate position in the queue according to the priority.

[0048] The method in this embodiment generates request control tasks and operation control tasks respectively based on control requests and equipment operation data, and schedules their execution based on preset priorities, thus solving the problem of fragmented control in existing microgrids. By actively identifying event-triggered data and abnormal data, high-priority control tasks can be generated in a timely manner, avoiding delays in safety-critical operations. The task queue mechanism ensures that important tasks are executed first, effectively addressing the response lag and scheduling chaos caused by the diverse types of microgrid equipment and the complexity of control, thereby improving the real-time performance, security, and coordination of control.

[0049] Figure 3 This is a schematic diagram illustrating the determination of a control task in a microgrid control method based on edge computing according to an embodiment of the present invention. The step of determining the operation control task corresponding to the equipment operation data may include: Step 701: Identify whether the device operation data includes event-triggered data that is not part of the regular data. Regular data refers to data collected from various devices in the microgrid according to a preset collection cycle. If event-triggered data is included, proceed to step S702; otherwise, proceed to step S703.

[0050] Regular data refers to data uploaded periodically by the device terminal during normal operation. The upload cycle for each type of regular data can be set differently, such as by minute, hour, or day. Event-triggered data is data generated after a special event occurs. These events can include: security protection events (such as fire alarm linkage, environmental changes, protection action events, equipment warning events, and data exceeding limits events), status change events (such as mode switching and device start / stop), and communication interaction events (such as communication failure, maintenance intervention, and permission changes).

[0051] Step S702: The event processing task corresponding to the event trigger data is designated as the operation control task. For example, when a protection action event occurs, the processing task for handling the subsequent operations of the protection action is designated as the operation control task.

[0052] Step S703: Identify whether there is any abnormal data in the equipment operation data.

[0053] Step S704: In the case of abnormal data, the analysis and response task for the abnormal data is taken as the operation control task.

[0054] Step S705: In the absence of abnormal data, generate operational control tasks based on regular data. Operational control tasks can be generated according to the control cycle of each control objective. One specific generation method is to obtain the pre-configured control cycles for each control requirement; generate control tasks to fulfill these control requirements according to the control cycles, which serve as operational control tasks. Thus, in the absence of triggering events or abnormal data, regular periodic tasks are executed. For example, if an energy storage device needs to optimize its control strategy daily, the daily energy storage control strategy optimization task is used as the operational control task to fulfill the energy storage device's control requirements.

[0055] The above process can avoid misjudgment of tasks caused by mixed data, ensure that event-triggered data is not missed and regular data is not over-interpreted, solve the problem of task and data scenario disconnect in traditional methods, adapt to the local processing characteristics of edge computing, reduce data transmission pressure and execution latency, fit the edge and cloud collaborative control architecture, and provide a foundation for subsequent task sorting and collaborative execution.

[0056] Figure 4This is a schematic diagram illustrating the execution of control tasks in a microgrid control method based on edge computing according to an embodiment of the present invention. The steps described above, in which the edge computing node and / or the cloud control platform of the microgrid sequentially execute the control tasks in the task execution queue, may include: Step S201: Obtain the control task at the head of the task execution queue as the target control task. The target control task may include simple control tasks implemented through logical judgment, complex control tasks that require the formulation of control strategies based on big data or deep learning, collaborative control tasks that require cross-regional collaboration, and data processing tasks involving data storage and analysis.

[0057] Step S202 involves breaking down the target control task into edge control subtasks and cloud control subtasks. Edge control subtasks are tasks that edge computing nodes can complete using their own data processing resources. These are typically tasks with high real-time requirements, low computational complexity, and rely solely on local data, such as generating control commands for a single device, message reporting, and data preprocessing. Cloud control subtasks are tasks that require the cloud control platform to access large amounts of data and models, and require massive storage capabilities. These are typically tasks involving global optimization goals, requiring access to historical big data or deep learning models, computational complexity exceeding a preset complexity limit, cross-regional collaborative decision-making, and long-term data storage, such as control strategy optimization, system collaborative control, log analysis, and model training.

[0058] In some embodiments, edge control subtasks may include: local control, protection action processing, data preprocessing, and human-computer interaction; cloud control subtasks may include: deep learning model training and runtime log analysis.

[0059] Step S203: Allocate data processing resources for edge computing nodes according to the data processing requirements of edge control subtasks, so that edge computing nodes can perform data processing for edge control subtasks.

[0060] Step S204: Determine the data transmission link based on the data transmission requirements of the cloud control subtask, and transmit the data required by the cloud control subtask to the cloud control platform via the data transmission link.

[0061] The microgrid control method based on edge computing in this embodiment dynamically divides the target control task into two sub-tasks: edge and cloud. It also matches the data processing resources and transmission link selection mechanism of the edge computing node, making full use of the data processing advantages of the edge computing node and the cloud, realizing the complementary synergy of cloud and edge capabilities, and meeting the high computing power and high real-time requirements of microgrid control.

[0062] Figure 5This is a schematic flowchart illustrating the allocation of data processing resources for edge computing nodes in a microgrid control method based on edge computing according to an embodiment of the present invention. The step of allocating data processing resources for edge computing nodes according to the data processing requirements of edge control subtasks may include: Step S301: Obtain the available computing power of the edge computing node to determine computing power constraints. The available computing power of the edge computing node includes resource indicators such as the total number of CPU cores and their utilization limit, and the total amount of memory.

[0063] Step S302 involves optimizing the computing power of each functional module of the edge computing node using a mixed-integer linear programming model, while satisfying computing power constraints. The mixed-integer linear programming model can calculate the baseline computing power of each functional module according to a preset ratio and dynamically adjust it based on the computing power requirements of the edge control subtasks. This dynamic adjustment process must be carried out under computing power constraints.

[0064] Figure 6 This is a schematic diagram of the functional modules of an edge computing node in a microgrid control method based on edge computing according to an embodiment of the present invention. The functional modules of the edge computing node 21 may include: a storage module 211, an alarm processing module 212, a control command generation module 213, a logic judgment module 214, and a log processing module 215.

[0065] The storage module 211 is used for localized data caching and management at the edge. The storage module 211 can temporarily store regular data and data triggered by sudden events collected periodically from microgrid devices, save local control strategy library, device configuration parameters and snapshots of edge computing node operating status, provide low-latency data read and write support for other modules, and cache data packets required by cloud control subtasks to be uploaded to the cloud.

[0066] The alarm processing module 212 is used to identify abnormal events and respond in real time. The alarm processing module 212 can identify event trigger data such as over-limit alarms and abnormal status uploaded by devices such as BMS and PCS, determine the abnormal handling plan, and generate abnormal handling control commands accordingly.

[0067] The control command generation module 213 is used to receive the decision results or the final control strategy after fusion from the logic judgment module and convert them into control commands that the controlled object device can recognize and process.

[0068] The logic judgment module 214 is used to perform threshold judgment and status logic judgment based on the device operation data and generate judgment results.

[0069] The log processing module 215 is used to collect and classify device operation logs, compress the logs, and upload key logs to the cloud according to the strategy.

[0070] The aforementioned functional modules rely on a Mixed Integer Linear Program (MILP) model for computing power allocation and cooperate to complete edge control sub-tasks. For example, when a sudden alarm event occurs, the alarm processing module 212 triggers the task, the logic judgment module 214 makes a logical decision, the control command generation module 213 issues the execution command, and the log processing module 215 and the storage function module 211 record relevant data. The MILP model can adjust the computing power ratio of each functional module according to the current edge control sub-task to ensure that critical tasks are processed first.

[0071] The process of computing power allocation in a MILP model is as follows: The baseline computing power of each module is initialized according to a preset ratio (Storage:Alarm:Control:Logic:Log = 2:3:2:2:1). For example, the initial ratio of computing power A of storage module 211, computing power B of alarm processing module 212, computing power C of control command generation module 213, computing power D of logic judgment module 214, and computing power E of log processing module 215 is: A:B:C:D:E = 2:3:2:2:1. Each module has a minimum computing power threshold (min). i (Minimum computing power to ensure basic functions) and maximum computing power threshold (max) i (Maximum computing power used to prevent single-module overload). Then define the continuous variable Δ. i (i=1~5) represents the incremental redundant computing power allocated to the i-th functional module; a binary variable y is introduced. i Setting it to 1 indicates that module i triggers high-priority resource preemption, while setting it to 0 indicates that module i does not trigger high-priority resource preemption; an integer variable k is introduced. i This represents the number of discrete resource blocks allocated to module i.

[0072] The objective function of the MILP model is designed to: under computing power constraints, make the actual computing power of each module as close as possible to the dynamic demand, while ensuring the supply of resources for high-priority tasks. The objective function transforms the absolute value constraint into a linear expression by introducing auxiliary variables to meet the MILP solution requirements. The objective function is solved using MILP, and the deviation between the actual resource utilization and the allocated value is monitored. This deviation is then used for the next round of model parameter tuning. For example, when event-triggered data occurs, the computing power demand of the alarm processing module 212 increases. The binary variable y2 of the alarm processing module 212 can be set to 1, and the MILP model automatically increases the continuous variable Δ2 of the alarm processing module 212, thus tilting computing power towards this module.

[0073] The microgrid control method based on edge computing in the above embodiments introduces a mixed integer linear programming model to dynamically optimize the computing power of edge computing nodes. Under the premise of meeting computing power constraints, it realizes precise resource allocation of modules such as storage, alarm, and control generation, avoiding resource crowding or idleness between modules caused by fixed allocation ratios, and meeting the computing power requirements under different control tasks.

[0074] Figure 7 This is a flowchart illustrating the process of determining a data transmission link in a microgrid control method based on edge computing according to an embodiment of the present invention. The steps of determining the data transmission link based on the data transmission requirements of the cloud control subtask include: Step S501: Determine the data transmission target of the cloud control subtask; Step S502: Traverse all network links to the data transmission target and obtain multiple evaluation indicators for each network link. The evaluation indicators include transmission bandwidth, transmission delay, delay jitter rate, and packet loss rate. Step S503: Use an evaluation function to fuse multiple evaluation indicators to obtain a comprehensive evaluation indicator; Step S504: Select the network link with the best comprehensive evaluation index as the data transmission link.

[0075] During the process of data being uploaded from edge computing node 21 to cloud control platform 24, or control policies being issued from cloud control platform 24 to edge computing node 21, the data will pass through at least one or more transmission nodes (such as switches or routers). Selecting the optimal data transmission link helps optimize data transmission efficiency and reduce packet loss. Simultaneously, during cloud-edge interaction, it helps ensure complete data upload and timely policy delivery.

[0076] In the above evaluation metrics, transmission bandwidth represents the maximum theoretical transmission rate of the link. Transmission delay is a combination of sending delay, propagation delay, processing delay, and queuing delay, where sending delay = data packet size / bandwidth, and propagation delay = link length / signal propagation speed. Transmission delay is generally required to be below 50ms. Latency jitter is the time difference between the delays of two adjacent data packets, generally required to be below 20ms. Packet loss rate is the percentage of lost packets out of the total number of packets sent.

[0077] The evaluation function can use a weighted fusion function, for example, it can be expressed as: Score = w1 × transmission bandwidth + w2 × (1 / transmission delay) + w3 × (1 - packet loss rate), where w1, w2, and w3 are the weights of transmission bandwidth, transmission delay, and packet loss rate, respectively. The transmission bandwidth, transmission delay, and packet loss rate mentioned above have been pre-normalized.

[0078] The microgrid control method based on edge computing in this embodiment constructs an evaluation function based on multi-dimensional network evaluation indicators to optimize link selection, so that the data transmission path selection shifts from dependence on a single indicator to multi-factor collaborative decision-making, effectively avoiding transmission interruption or strategy delay caused by link quality fluctuations, and enhancing the reliability of data upload for cloud control subtasks and the adaptive capability of cloud-edge communication links.

[0079] The steps described above for splitting the target control task into edge control subtasks and cloud control subtasks may include: dividing the target control task into edge control subtasks and cloud control subtasks according to preset task partitioning rules; wherein the task partitioning rules are based on the attributes of the subtasks, including real-time requirements, computational complexity, data dependency scope, and safety criticality level. The task partitioning rule library can predefine the matching rules between the attributes of each control task and the corresponding thresholds. For example, for real-time requirements: tasks with response times shorter than preset time limits (such as protection actions and emergency load adjustment) are classified as edge control subtasks. Another example is computational complexity: high-computing tasks requiring deep learning models or multi-timescale optimization are classified as cloud control subtasks. Yet another example is data dependency scope: tasks relying only on data from the local node are classified as edge control subtasks; tasks requiring cross-regional / historical big data analysis (such as full-site load prediction) are classified as cloud control subtasks. The task partitioning rules can be dynamically adjusted according to the current resource status and network conditions of the edge computing nodes, achieving environmental adaptability of the partitioning strategy.

[0080] After transmitting the data required for the cloud control subtask to the cloud control platform, the microgrid control method based on edge computing in this embodiment may further include: obtaining the cloud control strategy generated by the cloud control platform after executing the cloud control subtask; fusing the cloud control strategy with the local control strategy obtained by the edge computing node executing the edge control subtask to generate a final control command and issue it for execution. By fusing control strategies, the contradiction in traditional architectures where cloud strategies ignore local safety actions or edge strategies fail to consider global optimization goals is resolved. This allows the edge side to retain millisecond-level response capabilities to local emergencies while relying on continuous optimization based on network-wide data from the cloud, simultaneously achieving precise local response and continuous global optimization.

[0081] This embodiment also provides a computer program product 810, a computer-readable storage medium 820, and a computer device 830. Figure 8 This is a schematic diagram of a computer program product according to an embodiment of the present invention. Figure 9 This is a schematic diagram of a computer-readable storage medium according to an embodiment of the present invention. Figure 10 This is a schematic block diagram of a computer device according to an embodiment of the present invention.

[0082] Computer program product 810 includes computer program 811, which, when executed by processor 831, implements the steps of any of the above-described edge computing-based microgrid control methods. Computer-readable storage medium 820 stores the computer program 811 thereon, which, when executed by processor 831, implements the steps of any of the above-described edge computing-based microgrid control methods. Computer device 830 may include memory 832, processor 831, and computer program 811 stored in memory 832 and running on processor 831.

[0083] The computer program 811 used to perform the operations of this invention may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, status setting data, integrated circuit configuration data, or source code or object code written in any combination of one or more programming languages ​​and procedural programming languages.

[0084] Computer program 811 may execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the latter case, the remote computer may be connected to the user's computer via any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, to perform aspects of the invention, electronic circuits including, for example, programmable logic circuits, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs) may execute computer-readable program instructions to personalize the electronic circuits by utilizing state information of computer-readable program instructions.

[0085] For the purposes of this embodiment, computer program product 810 is a related product that includes computer program 811.

[0086] For the purposes of this embodiment, a computer-readable storage medium 820 is a tangible device capable of holding and storing a computer program 811. It can be any device capable of containing, storing, communicating, propagating, or transmitting the computer program 811 for use by or in conjunction with an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the computer-readable storage medium 820 include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable optical disc read-only memory (CD-ROM), digital versatile disc (DVD), memory stick, floppy disk, mechanical encoding device, and any suitable combination thereof.

[0087] Therefore, those skilled in the art should recognize that although numerous exemplary embodiments of the present invention have been shown and described in detail herein, many other variations or modifications conforming to the principles of the present invention can be directly determined or derived from the disclosure of the present invention without departing from the spirit and scope of the invention. Thus, the scope of the present invention should be understood and construed as covering all such other variations or modifications.

Claims

1. A microgrid control method based on edge computing, characterized in that... include: The edge computing nodes acquire equipment operation data and control requests from the microgrid. Determine the request control task corresponding to the control request; Determine the operation control tasks corresponding to the equipment operation data; The request control task and the operation control task are added to the task execution queue according to the pre-configured task priority based on the control strategy of the microgrid. The control tasks in the task execution queue are executed sequentially by the edge computing node and / or the cloud control platform of the microgrid.

2. The microgrid control method based on edge computing according to claim 1, characterized in that, The step of determining the operation control task corresponding to the equipment operation data includes: The system identifies whether the device operation data includes event-triggered data that is not part of the regular data. The regular data refers to data collected from each device in the microgrid according to a preset collection cycle. If so, the event processing task corresponding to the event triggering data shall be used as the operation control task; If not, identify whether there is abnormal data in the device operation data. If there is abnormal data, use the analysis and response task for the abnormal data as the operation control task. If there is no abnormal data, generate the operation control task based on the regular data.

3. The microgrid control method based on edge computing according to claim 2, characterized in that, The process of generating the operation control task based on conventional data also includes: Obtain the control cycle for each pre-configured control requirement; The control task is generated according to the control cycle to fulfill the control requirements, and serves as the operation control task.

4. The microgrid control method based on edge computing according to claim 1, characterized in that, The step of having the edge computing node and / or the cloud control platform of the microgrid sequentially execute the control tasks in the task execution queue includes: The control task at the head of the task execution queue is selected as the target control task. The target control task is broken down into an edge control sub-task and a cloud control sub-task. Data processing resources of edge computing nodes are allocated according to the data processing requirements of the edge control subtask, so that the edge computing nodes can perform the data processing of the edge control subtask. The data transmission link is determined based on the data transmission requirements of the cloud control subtask, and the data required by the cloud control subtask is transmitted to the cloud control platform via the data transmission link.

5. The microgrid control method based on edge computing according to claim 4, characterized in that, The step of allocating data processing resources for edge computing nodes according to the data processing requirements of the edge control subtask includes: Obtain the applicable computing power of the edge computing node to determine computing power constraints; By using a mixed-integer linear programming model to optimize the computing power occupied by each functional module of the edge computing node under the condition of satisfying the computing power constraints, the functional modules include: storage module, alarm processing module, control command generation module, logic judgment module, and log processing module.

6. The microgrid control method based on edge computing according to claim 4, characterized in that, The step of determining the data transmission link based on the data transmission requirements of the cloud control subtask includes: Determine the data transmission target of the cloud control subtask; Traverse all network links to the data transmission target and obtain multiple evaluation metrics for each network link, including transmission bandwidth, transmission delay, delay jitter rate, and packet loss rate. The multiple evaluation indicators are fused using an evaluation function to obtain a comprehensive evaluation indicator; The network link with the best comprehensive evaluation index is selected as the data transmission link.

7. The microgrid control method based on edge computing according to claim 4, characterized in that, The steps of breaking down the target control task into edge control sub-tasks and cloud control sub-tasks include: According to preset task division rules, the target control task is divided into edge control subtasks and cloud control subtasks; wherein The task division rules are based on the attributes of the subtasks, including real-time requirements, computational complexity, data dependency scope, and safety criticality level.

8. The microgrid control method based on edge computing according to claim 7, characterized in that, Edge control subtasks include: local control, protection action processing, data preprocessing, and human-computer interaction; The cloud-based control subtasks include: deep learning model training and runtime log analysis.

9. The microgrid control method based on edge computing according to claim 4, characterized in that, After transmitting the data required for the cloud control subtask to the cloud control platform, the following is also included: Obtain the cloud control strategy generated by the cloud control platform after executing the cloud control sub-task; The cloud-based control strategy is merged with the local control strategy obtained by the edge computing node executing the edge control sub-task to generate the final control command and issue it for execution.

10. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the steps of the microgrid control method based on edge computing as described in any one of claims 1 to 9.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that... When the computer program is executed by the processor, it implements the steps of the microgrid control method based on edge computing as described in any one of claims 1 to 9.

12. A computer device, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of the microgrid control method based on edge computing as described in any one of claims 1 to 9.

Images