A path constraint based edge collaborative task migration optimization method

Abstract

The present application relates to a kind of edge-end collaborative task migration optimization method based on path constraint, method is based on the realization of task migration optimization in the path constraint of edge-end collaborative process in edge intelligent computing framework, from theoretical level to the analysis of power internet of things edge intelligent operation cost optimization.Specifically include: using the edge intelligent framework of power internet of things, establish the migration system model under the edge computing environment;Based on the characteristics of each node in edge computing environment, the migration path is defined, and the different node migration task interval is obtained;Under the constraint condition of task migration path, more efficient edge-end collaborative task migration is realized by improved particle swarm algorithm.The method of the present application can be distinguished from the optimization of only energy consumption and task response time focused on migration in the traditional, using edge-end collaboration in the edge environment of power internet of things, the task migration optimization method is studied at theoretical level.

Description

Technical Field

[0001] This invention relates to a path-constrained edge-to-edge collaborative task migration optimization method, which enables the edge environment of the power Internet of Things (IoT) to optimize the time and energy consumption of the task migration process by restricting the task migration path through edge-to-edge collaboration. This realizes the implementation of edge computing to promote the ubiquitous power IoT and belongs to the field of ubiquitous power IoT. Background Technology

[0002] The goal of the ubiquitous power Internet of Things (IoT) is to achieve interconnectivity and human-machine interaction across all aspects of the power system, forming a smart service system characterized by comprehensive state perception, efficient information processing, and convenient and flexible applications. The State Grid Corporation of China (SGCC) has proposed a "chip-device-edge-network-cloud" overall architecture for the ubiquitous power IoT, where "edge" refers to building edge computing capabilities through the deployment of edge computing devices and distributed data centers. Currently, smart meters and networked distributed grid equipment have been upgraded and deployed, and can be considered as specific "terminals." With the introduction of edge computing technology, these scattered power devices will serve as important ports for the ubiquitous power IoT. Supported by next-generation information technology, they will achieve interconnection, human-machine interaction, comprehensive state perception, and efficient information processing, forming a powerful driving force for digital management and unlocking the value of big data. Of course, edge computing is not necessarily limited to meters and distribution equipment. The future of the Internet of Everything will enable every terminal within the IoT to have edge computing capabilities, thus constructing a grand digital landscape of cloud computing + edge computing.

[0003] Edge computing organically integrates computing, storage, and other resources at the network edge to build a unified user service platform. It responds promptly and processes task requests from network edge nodes according to the principle of proximity. Due to the limitations of edge node capabilities, resources, bandwidth, and energy, task migration is exceptionally important. This allows users to migrate computationally demanding tasks to idle edge servers. Compared to users accessing data in the cloud center, this reduces latency, decreases network load, and thus improves network performance and computing resource utilization.

[0004] Currently, some scholars both domestically and internationally have focused on research into network repair methods. However, the vast majority of these methods primarily concentrate on the energy consumption of migration and task response time, employing various genetic algorithms to improve their performance. Some studies simply divide tasks without considering the complex dependencies between them. Therefore, achieving efficient task migration in the edge environment of the power Internet of Things (IoT) through multi-entity collaboration is a pressing issue that needs to be addressed.

[0005] The State Grid Corporation of China (SGCC) has proposed a "chip-device-edge-network-cloud" overall architecture for the ubiquitous power Internet of Things (IoT). The "edge" refers to building edge computing capabilities through the deployment of edge computing devices and distributed data centers. In mobile edge computing, edge cloud servers are deployed at each base station, and network operators are responsible for forwarding and filtering data packets. The rise of task migration technology has introduced a new method to address the resource constraints of mobile terminals. How to achieve efficient task migration in the edge environment of the power IoT through collaboration between edge and endpoint devices is beneficial to enhancing the processing capacity and coverage of my country's power IoT, and thus has significant strategic importance for promoting the sound development of the State Grid and the power system. Summary of the Invention

[0006] To address the aforementioned technical shortcomings, the present invention aims to provide a path-constrained edge-to-edge collaborative task migration optimization method. By establishing a migration system model in an edge computing environment, constraining the task migration path based on the characteristics of each node in the environment, and utilizing an improved particle swarm optimization algorithm to optimize the task migration strategy, the method promotes the efficient operation of the power Internet of Things.

[0007] The technical solution adopted by this invention to achieve the above objectives is: a path-constrained edge-to-edge collaborative task migration optimization method, comprising the following steps:

[0008] Establish a task migration system model in the edge computing environment of the power Internet of Things;

[0009] Based on the task migration system model, determine the task migration strategy objectives and establish task migration path constraints;

[0010] Under the constraint of task migration path, the particle swarm optimization algorithm is used to optimize the task migration strategy in the edge computing environment of the power Internet of Things. The optimized task migration strategy that meets the objective of the task migration strategy is the optimized task migration strategy.

[0011] The optimized task migration strategy is used to realize task migration in the power Internet of Things edge computing environment.

[0012] A task migration system model is pre-established in the power Internet of Things edge computing environment, including:

[0013] As a node in the edge computing environment of the power Internet of Things, the mobile terminal divides the task queue into local tasks and transferable tasks. Local tasks are executed by the processing unit, while transferable tasks are optimized for migration strategy through particle swarm optimization. Based on the optimization result, the task is sent to the edge server or other mobile terminals that can communicate directly.

[0014] Edge servers execute portable tasks from mobile devices.

[0015] The objective of the task migration strategy is expressed as total energy consumption. :

[0016]

[0017] in, The total energy consumption of the task migration system model. This represents the set of all tasks that need to be executed in the power Internet of Things (IoT) edge computing environment. This represents the set of all nodes in an edge computing environment. This parameter represents the proportion of tasks that are migrated from task v to node n for execution. This represents the energy consumption of node n when executing task v. This indicates that when executing task v, it is necessary to use the node. and nodes Energy consumption for data transmission between them.

[0018] Energy consumption of node n when executing task v Obtain it through the following steps:

[0019]

[0020] This represents the workload of task v. This represents the amount of work performed by task v on node n. This represents the proportion of tasks v that are executed on node n;

[0021] set up Let be the CPU execution rate of node n, then the execution time of task v on node n is... for:

[0022]

[0023] set up Let be the execution power of node n, then the energy consumption of task v executing on node n. We obtain it from the following formula:

[0024] .

[0025] The optimization of task migration strategies in the power Internet of Things edge computing environment using the particle swarm optimization algorithm includes the following steps:

[0026] Step 1: Construct the migration matrix Each row represents the amount of work performed by a node across different tasks, and each column represents the amount of work performed by a task across different nodes.

[0027] Step 2: Set two initial values ​​that satisfy the task migration path constraints. and And respectively set them as the best historical particles in the particle swarm optimization algorithm. and population optimality Let each column of the migration matrix represent a particle in the particle swarm optimization algorithm, and let the migration matrix represent the population.

[0028] Step 3: For each iteration, select particles and populations that satisfy the task migration path constraints to update the historical particle optimal and population optimal; when the number of iterations reaches the set value, the optimization process stops, and the population optimal at this time is output, which is the task migration strategy.

[0029] The constraints of the task migration path include: task timing requirements, node load capacity, and task migration distance.

[0030] The specific conditions for satisfying the task migration path constraints are as follows:

[0031] Meets task timing requirements: Multiple tasks are executed in the set order;

[0032] Meeting the node's load capacity: The number of tasks on the node is less than the node's load limit;

[0033] The task migration distance requirement is met: the task migration distance is less than the threshold.

[0034] For each iteration, energy consumption calculations are performed between nodes of the task to calculate the total energy consumption. We obtain the following formula:

[0035]

[0036] in, This indicates that task v is in node and nodes Energy consumption between , These represent task v at node and nodes Energy consumption for executing subtasks; It is the energy transfer energy consumption coefficient. For nodes and nodes The path length between them.

[0037] The process of selecting particles and populations that satisfy the task migration path constraints and updating the historical particle optimal and population optimal includes the following steps:

[0038]

[0039]

[0040] in, Let be the flight velocity of particle i in the k-th iteration. It is used to adjust the inertia weight of the spatial search range. and It is a learning factor. and It is a random number. Denotes the transition matrix after the (k-1)th iteration. Column i;

[0041] The above formula yields a new transition matrix W after each iteration. The process of updating the transition matrix W is as follows:

[0042] The task migration path constraint condition is checked on the migration matrix W. If W satisfies the task migration path constraint condition, the next iteration process continues; otherwise, the next particle swarm optimization iteration process is directly entered.

[0043] Calculate the total energy consumption corresponding to the migration matrix W at this point based on the task migration strategy objective. With the current historical optimal particle and population optimality The corresponding total energy consumption is compared; if the current total energy consumption is smaller, then the current W replaces the historical particle with the highest current total energy consumption. or population optimal If the iteration continues, proceed to the next iteration; otherwise, proceed directly to the next iteration.

[0044] An edge-to-edge collaborative task migration optimization system based on path constraints includes:

[0045] The task migration initialization module is used to determine the task migration strategy objectives and establish task migration path constraints based on the task migration system model.

[0046] The task migration optimization module is used to optimize the task migration strategy in the power Internet of Things edge computing environment under the constraint of task migration path using the particle swarm algorithm. The optimized task migration strategy that meets the task migration strategy objective is the optimized task migration strategy.

[0047] The task migration module is used to migrate tasks in the power Internet of Things edge computing environment according to the optimized task migration strategy.

[0048] The task migration system model includes:

[0049] Mobile terminals, as nodes in the power Internet of Things edge computing environment, are used to divide the task queue into local tasks and transferable tasks. Local tasks are executed by the current mobile terminal, while transferable tasks are optimized for migration strategies through particle swarm optimization, so that the tasks can be sent to edge servers or other mobile terminals that can communicate directly based on the optimization results.

[0050] Edge servers are used to perform portable tasks from mobile devices.

[0051] The present invention has the following beneficial effects and advantages:

[0052] 1. This invention establishes a migration system model in an edge computing environment, providing a tool for the application of edge computing in the ubiquitous power Internet of Things.

[0053] 2. This invention differs from traditional optimization methods that focus solely on energy consumption and task response time during migration. By analyzing the characteristics of each node in the edge computing environment and constraining the task migration path, it helps improve the operational efficiency and stability of the power Internet of Things.

[0054] 3. To meet the application background of the power Internet of Things, this invention introduces an improved discrete binary particle swarm optimization algorithm to optimize the task migration strategy. Attached Figure Description

[0055] Figure 1 This is a flowchart of the edge-to-edge collaborative task migration optimization method;

[0056] Figure 2 This is a schematic diagram of the migration system model;

[0057] Figure 3 This is a schematic diagram of the task partitioning model;

[0058] Figure 4 This is a flowchart of the particle swarm optimization algorithm under path constraints. Detailed Implementation

[0059] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments.

[0060] This invention proposes an edge-to-edge collaborative task migration optimization method based on path constraints. By modeling the task migration system in the edge computing environment, it differs from traditional optimization methods that only focus on the energy consumption and task response time of migration. By analyzing the characteristics of each node in the edge computing environment, it constrains the task migration path and uses an improved discrete binary particle swarm optimization algorithm to optimize the task migration strategy.

[0061] Establish a task migration system model in the edge computing environment of the power Internet of Things;

[0062] Based on the application scenarios of edge computing in the ubiquitous power Internet of Things, the objectives of the task migration strategy are defined;

[0063] A task migration path constraint method is established from three aspects: task timing requirements, node load capacity, and task migration distance.

[0064] Under the constraint of task migration path, the particle swarm optimization algorithm is used to optimize the task migration strategy in the edge computing environment, and the optimized task migration strategy is obtained.

[0065] Based on the application background of task migration in the edge environment of the power Internet of Things, the migration matrix, particles and population in the particle swarm algorithm are defined.

[0066] Based on the task migration path constraints, the particle swarm optimization algorithm is used for optimization.

[0067] Based on the pre-set maximum number of iterations, the particle swarm optimization steps are repeatedly executed to complete the task migration in the edge environment of the power Internet of Things.

[0068] The establishment of the task migration system model in the edge computing environment of the power Internet of Things is specifically as follows:

[0069] The model's objects are divided into two parts: mobile terminals and edge servers. The mobile terminals include a computing task queue, processing units, and a transmission unit. The edge servers include processing units and server waiting queues, such as... Figure 2 As shown, in the terminal, the task queue is divided into locally executed tasks and transferable tasks based on the characteristics of the system input tasks. Locally executed tasks are directly executed by the local processing unit, while transferable tasks are uploaded by the sending unit to the edge server or other directly communicable mobile terminals, where they are executed by the edge server and mobile terminal processing units. Finally, after execution, the results are fed back to the mobile terminal. In the migration system model, it is assumed that the computing power of the edge server is far greater than that of the mobile terminal.

[0070] The goal of the task migration optimization is to reduce the total system energy consumption, specifically expressed as follows: The research aims to reduce migration energy consumption and time complexity. By optimizing the time and energy consumption model, the system is divided into two parts: the first part represents the energy consumption of tasks executed locally and at the edge, and the second part represents the energy consumption of data transmission between dependent tasks.

[0071] The task migration path constraint method described above takes into account three constraints: task timing requirements, node load capacity, and task migration distance.

[0072] In this context, the task migration path constraints in these three aspects will be applied to the subsequent optimization process.

[0073] The optimization of the task migration strategy using the particle swarm optimization algorithm is specifically as follows:

[0074] Step 1: Construct the migration matrix Each row represents the amount of work performed by a node across different tasks, and each column represents the amount of work performed by a task across different nodes.

[0075] Step 2: Based on the constraints in the task migration path constraint method, two initial values ​​that meet the conditions can be set, and these values ​​can be set as the historical particle optimal and the population optimal in the particle swarm algorithm, respectively. Let each column in the migration matrix be a particle in the particle swarm algorithm.

[0076] Step 3: Each iteration of the particles and population needs to undergo constraint checks in the task migration path constraint method. Only when the constraints in the task migration path constraint method are met can the historical particle optimal and population optimal be updated.

[0077] By following the steps above, the task migration strategy can be optimized.

[0078] The particle swarm optimization process, which uses the particle swarm optimization algorithm to optimize the task migration strategy, is iterated to obtain the final optimized task migration strategy. The iterative process of the particle swarm optimization algorithm requires pre-setting the maximum number of iterations and the energy transfer energy consumption coefficient, specifically as follows:

[0079] Step 1: When the number of iterations is reached, the optimization process stops, and the population optimum at this point is output.

[0080] Step 2, Energy Transfer Energy Consumption Coefficient ,when The larger the value, the fewer migration tasks will be in the entire strategy.

[0081] By following the steps above, more efficient edge-to-device collaborative task migration can be achieved while meeting the application requirements of the power Internet of Things (IoT) edge environment, thereby promoting the implementation of ubiquitous power IoT through edge computing.

[0082] This invention provides a path-constrained edge-to-edge collaborative task migration optimization method to improve the operational efficiency and stability of the power Internet of Things.

[0083] like Figure 1 As shown, the specific steps include:

[0084] Step 1: Model the migration system in the edge computing environment. The model objects are divided into two parts: mobile terminals and edge servers. The mobile terminals include computing task queues, processing units, and transmission units. The edge servers include processing units and server waiting queues, such as... Figure 2 As shown, in the terminal, the task queue is divided into locally executed tasks and transferable tasks based on the characteristics of the system input tasks. Locally executed tasks are directly executed by the local processing unit, while transferable tasks are uploaded by the sending unit to the edge server or other directly communicable mobile terminals, where they are executed by the edge server and mobile terminal processing units. Finally, after execution, the results are fed back to the mobile terminal. In the migration system model, it is assumed that the computing power of the edge server is far greater than that of the mobile terminal.

[0085] Step 2: Define the task migration strategy objectives based on the application scenarios of edge computing in the ubiquitous power Internet of Things. The research objective is to reduce migration energy consumption and time complexity. Figure 2 As can be seen, the tasks of the mobile terminal are divided into locally executed and portable tasks. This patent uses variables to represent the task division and the dependencies between tasks, where the subscript represents the task. By optimizing the time and energy consumption model, the system is divided into two parts: the first part represents the energy consumption of tasks executed locally and at the edge, and the second part represents the energy consumption of data transmission between dependent tasks. The total system energy consumption is expressed as:

[0086]

[0087] This transforms the system model into an unrestricted planning problem, ultimately yielding a set of optimal migration decisions to determine the migration locations for subtasks. This represents the set of all tasks that need to be executed in this edge computing environment. Represents the set of all nodes in the edge environment. This parameter represents the proportion of tasks that are migrated from task v to node n for execution. This represents the energy consumption of node n when executing task v. This indicates that when executing task v, it is necessary to use the node. and nodes Energy consumption for data transmission between them.

[0088] Step 3: Based on the settings in Step 2, establish a task migration path constraint method. This invention considers three constraints during task migration path planning: task timing requirements, node load capacity, and task migration distance. The specific task partitioning model can be found here. Figure 3 Based on the order in which nodes receive tasks to be executed, a directed acyclic graph is constructed, where the direction of the edges indicates the direction in which tasks can migrate. For example... Figure 3 To execute task 2, tasks 1 and j must be completed first, and task j can migrate tasks to nodes 2 and i. The migration path constraint process involves three aspects: 1. Task order; 2. Node load; 3. Migration path distance.

[0089] 1. Task order: such as Figure 3 In the example, the direction of the edges represents the order of tasks. To execute task 2 or task c, task 1 must be completed first; and to execute task d, tasks 2 and c must be completed first. This constraint is an objective condition that edge computing must adhere to in the application of ubiquitous power Internet of Things.

[0090] 2. Node load: Each node has an initial load limit. The total number of tasks processed by a node at any given time cannot exceed this upper limit. Furthermore, the remaining load of each node should be considered during the task distribution node selection process; a higher remaining load corresponds to a larger number of tasks that should be distributed.

[0091] 3. Migration path distance: In edge environments, considering the energy and time consumption of transfer between task nodes, the amount of tasks transmitted to neighboring nodes at secondary or even greater distances will be constrained by distance. In this invention, the larger the migration path distance, the smaller the proportion of tasks distributed will be.

[0092] In this context, the task migration path constraints in these three aspects will be applied to the subsequent optimization process.

[0093] Step four, following step three, this invention proposes to use a particle swarm optimization algorithm to optimize the task migration strategy. The total system energy consumption is the fitness function value in this invention. Let the number of all tasks to be executed in the edge computing environment be... The total number of nodes in the edge environment is Then the problem can be regarded as a question about a The transition matrix is ​​optimized and can be expressed as:

[0094]

[0095] in This represents the j-th task. At the node The amount of work performed reveals the following relationship:

[0096]

[0097] This represents the workload of task v. This represents the amount of work performed by task v on node n. This represents the proportion of tasks v that are executed on node n.

[0098] Set as Given the CPU execution speed of node n, the execution time of task v on node n is:

[0099]

[0100] set up Given the execution power of node n (which is set to a constant in this invention), the energy consumption of task v executing on node n is... We obtain it from the following formula:

[0101]

[0102] In addition to the parameters mentioned above, this invention also calculates the energy consumption between nodes of the task:

[0103]

[0104] in It is the energy transfer energy consumption coefficient, with a value between (1,2). For nodes and nodes The path length between them.

[0105] Based on the constraints in step three, two initial values ​​that meet the conditions can be set. and And respectively set them as the best historical particles in the particle swarm optimization algorithm. and population optimality For the transition matrix Let each column Let be a particle in the particle swarm optimization algorithm. Then, the speed update of the particle swarm optimization process in this invention is as follows:

[0106]

[0107] in Let be the flight velocity of particle i in the k-th iteration. It is a non-negative number used to adjust the spatial search range, representing the inertial weight, which is set to... , and It is the learning factor (all are set to 2 in this invention). and Given two random numbers, whose values ​​range from [0,1]. Denotes the transition matrix after the (k-1)th iteration. Column i.

[0108] The position update in the particle swarm optimization process of this invention is as follows:

[0109]

[0110] The specific optimization process of the particle swarm optimization algorithm is as follows: Figure 4As shown, it is worth noting that the particles and population obtained in each iteration need to undergo the constraint check in step three. Only by satisfying the constraints in step three can the historical particles be optimally matched. and population optimality Update.

[0111] Step 5: Following steps 3 and 4, this invention obtains the final optimized edge environment task migration strategy after setting the initial values. Specific initial values ​​include the maximum number of iterations K. When the number of iterations reaches K, the optimization process stops, and the population optimum at that point is output. Energy transfer energy consumption coefficient ,when A larger value results in fewer migration tasks in the overall strategy; this invention sets it to 1.2. If there are time constraints in the actual execution of tasks, the time consumption of each task during execution can be added in step three.

[0112] Through this step, we will finally obtain the edge-to-edge collaborative task migration strategy in the power Internet of Things edge environment.

[0113] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of embodiments combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0114] This invention is described with reference to a process flowchart according to embodiments of this application. It should be understood that each step in the flowchart can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing device, generate instructions for implementing the process. Figure 1 A device for a function specified in one or more processes.

[0115] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 The function specified in one or more processes.

[0116] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 Steps of a specified function in one or more processes.

[0117] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for optimizing the migration of edge-to-edge collaborative tasks based on path constraints, characterized in that, Includes the following steps: Establish a task migration system model in the edge computing environment of the power Internet of Things; Based on the task migration system model, determine the task migration strategy objectives and establish task migration path constraints; Under the constraint of task migration path, the particle swarm optimization algorithm is used to optimize the task migration strategy in the edge computing environment of the power Internet of Things. The optimized task migration strategy that meets the objective of the task migration strategy is the optimized task migration strategy. Implement task migration in the power Internet of Things edge computing environment based on the optimized task migration strategy; The objective of the task migration strategy is expressed as total energy consumption. : ； in, The total energy consumption of the task migration system model. This represents the set of all tasks that need to be executed in the power Internet of Things (IoT) edge computing environment. This represents the set of all nodes in an edge computing environment. This parameter represents the proportion of tasks that are migrated from task v to node n for execution. This represents the energy consumption of node n when executing task v. This indicates that when executing task v, it is necessary to use the node. and nodes Energy consumption for data transmission between them.

2. The edge-to-edge collaborative task migration optimization method based on path constraints according to claim 1, characterized in that, A task migration system model is pre-established in the power Internet of Things edge computing environment, including: As a node in the edge computing environment of the power Internet of Things, the mobile terminal divides the task queue into local tasks and transferable tasks. Local tasks are executed by the processing unit, while transferable tasks are optimized for migration strategy through particle swarm optimization. Based on the optimization result, the task is sent to the edge server or other mobile terminals that can communicate directly. Edge servers execute portable tasks from mobile devices.

3. The edge-to-edge collaborative task migration optimization method based on path constraints according to claim 1, characterized in that, Energy consumption of node n when executing task v Obtain it through the following steps: ； This represents the workload of task v. This represents the amount of work performed by task v on node n. This represents the proportion of tasks v that are executed on node n; set up Let be the CPU execution rate of node n, then the execution time of task v on node n is... for: ； set up Let be the execution power of node n, then the energy consumption of task v executing on node n. We obtain it from the following formula: 。 4. The edge-to-edge collaborative task migration optimization method based on path constraints according to claim 1, characterized in that, The optimization of task migration strategies in the power Internet of Things edge computing environment using the particle swarm optimization algorithm includes the following steps: Step 1: Construct the migration matrix Each row represents the amount of work performed by a node across different tasks, and each column represents the amount of work performed by a task across different nodes. Step 2: Set two initial values ​​that satisfy the task migration path constraints. and And respectively set them as the best historical particles in the particle swarm optimization algorithm. and population optimality Let each column of the migration matrix represent a particle in the particle swarm optimization algorithm, and let the migration matrix represent the population. Step 3: For each iteration, select particles and populations that satisfy the task migration path constraints to update the historical particle optimal and population optimal; when the number of iterations reaches the set value, the optimization process stops, and the population optimal at this time is output, which is the task migration strategy.

5. A path-constrained edge-to-edge collaborative task migration optimization method according to claim 1 or 4, characterized in that, The constraints of the task migration path include: task timing requirements, node load capacity, and task migration distance. The specific conditions for satisfying the task migration path constraints are as follows: Meets task timing requirements: Multiple tasks are executed in the set order; Meeting the node's load capacity: The number of tasks on the node is less than the node's load limit; The task migration distance requirement is met: the task migration distance is less than the threshold.

6. The edge-to-edge collaborative task migration optimization method based on path constraints according to claim 4, characterized in that, For each iteration, energy consumption calculations are performed between nodes of the task to calculate the total energy consumption. We obtain the following formula: ； in, This indicates that task v is in node and nodes Energy consumption between , These represent task v at node and nodes Energy consumption for performing tasks; It is the energy transfer energy consumption coefficient, with a value between (1,2). For nodes and nodes The path length between them.

7. The edge-to-edge collaborative task migration optimization method based on path constraints according to claim 4, characterized in that, The process of selecting particles and populations that satisfy the task migration path constraints and updating the historical particle optimal and population optimal includes the following steps: ； ； in, Let be the flight velocity of particle i in the k-th iteration. It is used to adjust the inertia weight of the spatial search range. and It is a learning factor. and It is a random number. Denotes the transition matrix after the (k-1)th iteration. Column i, This represents the current total energy consumption of the particles; The above formula yields a new transition matrix W after each iteration. The process of updating the transition matrix W is as follows: The task migration path constraint condition is checked on the migration matrix W. If W satisfies the task migration path constraint condition, the next iteration process continues; otherwise, the next particle swarm optimization iteration process is directly entered. Calculate the total energy consumption corresponding to the migration matrix W at this point based on the task migration strategy objective. With the current historical optimal particle and population optimality The corresponding total energy consumption is compared; if the current total energy consumption is smaller, then the current W replaces the historical particle with the highest current total energy consumption. or population optimal If the iteration continues, proceed to the next iteration; otherwise, proceed directly to the next iteration.

8. A path-constrained edge-end collaborative task migration optimization system, the system being used to implement the path-constrained edge-end collaborative task migration optimization method as described in any one of claims 1-7, characterized in that, include: The task migration initialization module is used to determine the task migration strategy objectives and establish task migration path constraints based on the task migration system model. The task migration optimization module is used to optimize the task migration strategy in the power Internet of Things edge computing environment under the constraint of task migration path using the particle swarm algorithm. The optimized task migration strategy that meets the task migration strategy objective is the optimized task migration strategy. The task migration module is used to migrate tasks in the power Internet of Things edge computing environment according to the optimized task migration strategy.

9. The edge-to-edge collaborative task migration optimization system based on path constraints according to claim 8, characterized in that, The task migration system model includes: Mobile terminals, as nodes in the power Internet of Things edge computing environment, are used to divide the task queue into local tasks and transferable tasks. Local tasks are executed by the current mobile terminal, while transferable tasks are optimized for migration strategies through particle swarm optimization, so that the tasks can be sent to edge servers or other mobile terminals that can communicate directly based on the optimization results. Edge servers are used to perform portable tasks from mobile devices.

Images