A computing power optimization method, storage medium, and device based on an industrial control gateway

Abstract

This invention discloses a computing power optimization method based on an industrial control gateway, characterized by the following steps: collecting computing power index data of each node through the industrial control gateway and calculating the comprehensive computing power score of each node; setting the number of grids and setting the upper limit of the total comprehensive computing power score for each grid; dividing all nodes into corresponding grids using the K-means algorithm; calculating the comprehensive performance evaluation value of each grid based on the computing task completion time, resource utilization, and communication latency; allocating computing tasks to the grid with the lowest comprehensive performance evaluation value; and executing computing tasks through the nodes in the grid. This invention, through grid partitioning, achieves refined management of computing resources and a high degree of matching between computing task execution and performance, improving production efficiency and reducing operation and maintenance costs.

Description

Technical Field

[0001] This invention belongs to the field of computing power optimization technology, specifically, it relates to a computing power optimization method, storage medium and device based on an industrial control gateway. Background Technology

[0002] In distributed computing environments, dynamically allocating tasks allows idle nodes to utilize their computing power for other tasks, preventing resource waste and improving resource utilization. Assigning tasks to nodes with matching computing power characteristics significantly shortens task completion time and improves real-time performance. Furthermore, when a node fails, quickly migrating its tasks to other healthy nodes avoids task interruption and enhances reliability. Therefore, in fields such as intelligent manufacturing and smart cities, computing power optimization has become a key technology for ensuring high-performance and highly reliable system operation, directly driving enterprise digital transformation and intelligent upgrading.

[0003] However, existing computing power optimization methods still have the following shortcomings: Existing computing power optimization methods rely on static resource allocation and fixed scheduling strategies. However, in actual operating environments, node computing power may change due to factors such as hardware aging, temperature changes, and interference from concurrent tasks. Existing computing power optimization methods are difficult to respond to node computing power fluctuations in real time, resulting in local resource idleness or global task accumulation, thereby reducing resource utilization and increasing communication latency. Summary of the Invention

[0004] To address the problems existing in the prior art, this invention provides a computing power optimization method, storage medium, and device based on an industrial control gateway. By dynamically adjusting the grid through node computing power fluctuations, resource utilization is improved and communication latency is greatly reduced.

[0005] To achieve the above technical objectives, the present invention adopts the following technical solution:

[0006] A computing power optimization method based on an industrial control gateway includes the following steps: Step S1: Collect computing power index data of each node through the industrial control gateway and calculate the comprehensive computing power score of each node; Step S2: Set the number of grids and set the upper limit of the total comprehensive computing power score for each grid. Use the K-means algorithm to divide all nodes into the corresponding grids. Step S3: Calculate the comprehensive performance evaluation value of each grid based on the completion time of the computing task, resource utilization rate and communication latency, allocate the computing task to the grid with the smallest comprehensive performance evaluation value, and execute the computing task through the nodes in the grid.

[0007] Furthermore, the computing power metrics include: node CPU utilization, node memory usage, and node GPU utilization.

[0008] Furthermore, the calculation process for the comprehensive computing power score of each node is as follows:

[0009] in, Indicates the first The overall computing power score of each node. Indicates the first CPU utilization of each node Indicates the first CPU capacity of each node This indicates the CPU weight determined by the task type; Indicates the first Memory usage of each node, Indicates the first The memory capacity of each node, This represents the memory weight determined by the task type; Indicates the first GPU utilization of each node, Indicates the first GPU capacity per node This represents the GPU weights determined by the task type.

[0010] Further, step S2 includes the following sub-steps: Step S2.1: Set the number of grid cells to [value]. Set the upper limit of the total comprehensive computing power score for each grid, and randomly select... The comprehensive computing power score of each node serves as the initialization center for the corresponding grid; Step S2.2: Calculate the overall computing power score of each node and the Euclidean distance to the center of each grid, and assign the node to the grid with the shortest Euclidean distance; Step S2.3: Determine whether the total comprehensive computing power score of each grid is less than the upper limit of the total comprehensive computing power score of the corresponding grid. If yes, proceed to step S2.4; otherwise, arrange the nodes in the grid in descending order of Euclidean distance, and migrate the nodes with larger Euclidean distances to the grids with the second shortest distance from the grid center, until the total comprehensive computing power score of each grid is less than the upper limit of the total comprehensive computing power score of the corresponding grid. Step S2.4: Calculate the average comprehensive computing power score of nodes in each grid, update the corresponding grid center, and repeat steps S2.2-S2.3 until the grid center no longer changes, and assign all nodes to the corresponding grid.

[0011] Furthermore, the calculation process for the comprehensive performance evaluation value of each grid in step S3 is as follows:

[0012] in, Indicates the first The overall performance evaluation value of each grid, Indicates the computing task in the th The completion time of each grid. This represents the shortest completion time for a computing task on the grid. This indicates the maximum completion time of a computing task on the grid. Weighting coefficients representing the completion time of computing tasks; Indicates the computing task in the th Resource utilization rate of each grid This represents the minimum resource utilization rate of a computing task on the grid. This represents the maximum resource utilization rate of computing tasks on the grid. Weighting coefficients representing the utilization rate of computing power tasks; Indicates the computing task in the th Communication latency of each grid This represents the minimum communication latency of a computing task on the grid. This represents the maximum communication latency of the computing task on the grid.

[0013] Furthermore, it also includes: feeding back the real-time computing power index data of the nodes to the industrial control gateway, calculating the total comprehensive computing power score of all nodes in each grid, and if the total comprehensive computing power score of each grid is above the upper limit, dividing all nodes into the corresponding grids using the K-means algorithm.

[0014] Furthermore, the present invention also provides a computer-readable storage medium storing a computer program that enables a computer to execute the aforementioned computing power optimization method based on an industrial control gateway.

[0015] Furthermore, the present invention also provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements the aforementioned computing power optimization method based on an industrial control gateway.

[0016] Compared with existing technologies, the present invention has the following beneficial effects: The computing power optimization method based on industrial control gateways collects multi-dimensional computing power index data from each node through the industrial control gateway, avoiding the one-sidedness of evaluating a single computing power index. It reflects the actual computing power bottleneck of each node through comprehensive computing power scoring, providing an accurate basis for subsequent grid partitioning. Simultaneously, the present invention uses the K-means algorithm to partition all nodes into corresponding grids, realizing pooled management of computing power resources. Each grid is treated as an independent scheduling unit, dynamically balancing the load of nodes in the grid and preventing local overload or idleness. Furthermore, the present invention calculates the comprehensive performance evaluation value of each grid based on computing task completion time, resource utilization, and communication latency, allocating computing tasks to the grid with the lowest comprehensive performance evaluation value. The comprehensive performance evaluation value of the grid comprehensively measures the efficiency of the grid in executing computing tasks, achieving reasonable allocation of computing power tasks. At the same time, the same computing task is executed on nodes within the same grid, reducing cross-grid communication requirements and lowering network bandwidth consumption and latency. Attached Figure Description

[0017] Figure 1 This is a flowchart of the computing power optimization method based on an industrial control gateway according to the present invention; Figure 2 This is a flowchart illustrating how the K-means algorithm is used to divide all nodes into corresponding grids in this invention. Detailed Implementation

[0018] The technical solution of the present invention will be further explained and described below with reference to the accompanying drawings.

[0019] like Figure 1 This is a flowchart of the computing power optimization method based on an industrial control gateway according to the present invention. The computing power optimization method includes the following steps: Step S1: Collect computing power index data of each node through the industrial control gateway, and calculate the comprehensive computing power score of each node. Specifically, the computing power index data includes the node's CPU utilization rate, the node's memory occupancy rate, and the node's GPU utilization rate. This avoids the one-sidedness of evaluating a single computing power index. The comprehensive computing power score reflects the actual computing power bottleneck of each node, providing an accurate basis for subsequent grid partitioning.

[0020] The calculation process for the comprehensive computing power score of each node is as follows:

[0021] in, Indicates the first The overall computing power score of each node. Indicates the first CPU utilization of each node Indicates the first CPU capacity of each node This indicates the CPU weight determined by the task type; Indicates the first Memory usage of each node, Indicates the first The memory capacity of each node, This represents the memory weight determined by the task type; Indicates the first GPU utilization of each node, Indicates the first GPU capacity per node This represents the GPU weights determined by the task type.

[0022] Step S2: Set the number of grids and the upper limit of the total comprehensive computing power score for each grid. Using the K-means algorithm, all nodes are divided into corresponding grids, achieving pooled management of computing resources. Each grid is treated as an independent scheduling unit, dynamically balancing the load of nodes within the grid and preventing local overload or idleness. For example... Figure 2 This includes the following sub-steps: Step S2.1: Set the number of grid cells to [value]. Set the upper limit of the total comprehensive computing power score for each grid, and randomly select... The comprehensive computing power score of each node serves as the initialization center for the corresponding grid; Step S2.2: Calculate the overall computing power score of each node and the Euclidean distance to the center of each grid, and assign the node to the grid with the shortest Euclidean distance; Step S2.3: Determine whether the total comprehensive computing power score of each grid is less than the upper limit of the total comprehensive computing power score of the corresponding grid. If yes, proceed to step S2.4; otherwise, arrange the nodes in the grid in descending order of Euclidean distance, and migrate the nodes with larger Euclidean distances to the grids with the second shortest distance from the grid center, until the total comprehensive computing power score of each grid is less than the upper limit of the total comprehensive computing power score of the corresponding grid. This can achieve dynamic resource balancing of the grid, avoid grid overload or idleness, and at the same time, by prioritizing the removal of nodes with lower contributions to the grid, the frequency of node movement across grids can be reduced.

[0023] Step S2.4: Calculate the average comprehensive computing power score of nodes in each grid, update the corresponding grid center, and repeat steps S2.2-S2.3 until the grid center no longer changes, and assign all nodes to the corresponding grid.

[0024] Step S3: Calculate the comprehensive performance evaluation value for each grid based on the computing task completion time, resource utilization, and communication latency. Assign computing tasks to the grid with the lowest comprehensive performance evaluation value, and execute the tasks through nodes within that grid. The computing task completion time reflects the grid's speed in processing computing tasks, resource utilization reflects the grid's effective use of resources, and communication latency measures the efficiency of data transmission within the grid. The comprehensive performance evaluation value of each grid comprehensively measures its processing capacity, resource utilization, and data transmission performance, achieving reasonable allocation of computing tasks. Simultaneously, the same computing task is executed on nodes within the same grid, reducing cross-grid communication needs, lowering network bandwidth consumption and latency, and meeting the grid performance evaluation requirements in different scenarios.

[0025] The calculation process for the comprehensive performance evaluation value of each grid in this invention is as follows:

[0026] in, Indicates the first The overall performance evaluation value of each grid, Indicates the computing task in the th The completion time of each grid. This represents the shortest completion time for a computing task on the grid. This indicates the maximum completion time of a computing task on the grid. Weighting coefficients representing the completion time of computing tasks; Indicates the computing task in the th Resource utilization rate of each grid This represents the minimum resource utilization rate of a computing task on the grid. This represents the maximum resource utilization rate of computing tasks on the grid. Weighting coefficients representing the utilization rate of computing power tasks; Indicates the computing task in the th Communication latency of each grid This represents the minimum communication latency of a computing task on the grid. This represents the maximum communication latency of the computing task on the grid.

[0027] In one technical solution of the present invention, the method further includes: feeding back the real-time computing power index data of the nodes to the industrial control gateway, which can accurately grasp the actual working capacity of each node; calculating the total comprehensive computing power score of all nodes in each grid; if the total comprehensive computing power score of each grid is above the upper limit, dividing all nodes into corresponding grids using the K-means algorithm; and dynamically adjusting the grids based on the feedback of the total comprehensive computing power score of all nodes in the grid, transferring the computing power load from high-load grids to low-load grids, achieving computing power balance between grids, avoiding the backlog of computing power tasks in some grids due to insufficient computing power, and improving the overall resource utilization rate of computing power tasks.

[0028] This invention provides a computing power optimization method based on industrial control gateways. By monitoring the computing power data indicators of nodes in real time through industrial control gateways and performing grid partitioning, it achieves refined management of computing power resources and a high degree of matching between computing power task execution, thereby improving production efficiency and reducing operation and maintenance costs.

[0029] In one technical solution of the present invention, a computer-readable storage medium is also provided, storing a computer program that enables a computer to execute the computing power optimization method based on an industrial control gateway of the present invention.

[0030] In one technical solution of the present invention, an electronic device is also provided, including: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the computing power optimization method based on an industrial control gateway of the present invention.

[0031] In the embodiments disclosed in this application, a computer storage medium may be a tangible medium that may contain or store programs for use by or in conjunction with an instruction execution system, apparatus, or device. The computer storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of computer storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, and portable compact disc read-only memory (CD). ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0032] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this application can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0033] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A computing power optimization method based on an industrial control gateway, characterized in that, Includes the following steps: Step S1: Collect computing power index data of each node through the industrial control gateway and calculate the comprehensive computing power score of each node; Step S2: Set the number of grids and set the upper limit of the total comprehensive computing power score for each grid. Use the K-means algorithm to divide all nodes into the corresponding grids. Step S3: Calculate the comprehensive performance evaluation value of each grid based on the completion time of the computing task, resource utilization rate and communication latency, allocate the computing task to the grid with the smallest comprehensive performance evaluation value, and execute the computing task through the nodes in the grid.

2. The computing power optimization method based on an industrial control gateway according to claim 1, characterized in that, The computing power metrics include: node CPU utilization, node memory usage, and node GPU utilization.

3. The computing power optimization method based on an industrial control gateway according to claim 2, characterized in that, The calculation process for the comprehensive computing power score of each node is as follows: in, Indicates the first The overall computing power score of each node. Indicates the first CPU utilization of each node Indicates the first CPU capacity of each node This indicates the CPU weight determined by the task type; Indicates the first Memory usage of each node, Indicates the first The memory capacity of each node, This represents the memory weight determined by the task type; Indicates the first GPU utilization of each node, Indicates the first GPU capacity per node This represents the GPU weights determined by the task type.

4. The computing power optimization method based on an industrial control gateway according to claim 1, characterized in that, Step S2 includes the following sub-steps: Step S2.1: Set the number of grid cells to [value]. Set the upper limit of the total comprehensive computing power score for each grid, and randomly select... The overall computing power score of each node serves as the initialization center for the grid; Step S2.2: Calculate the overall computing power score of each node and the Euclidean distance to the center of each grid, and assign the node to the grid with the shortest Euclidean distance; Step S2.3: Determine whether the total comprehensive computing power score of each grid is less than the upper limit of the total comprehensive computing power score of the corresponding grid. If so, proceed to step S2.

4. Otherwise, the nodes in the grid are arranged in descending order of Euclidean distance, and the nodes with larger Euclidean distances are migrated to the grid with the second shortest distance from the grid center, until the total comprehensive computing power score of each grid is less than the upper limit of the total comprehensive computing power score of the corresponding grid. Step S2.4: Calculate the average comprehensive computing power score of nodes in each grid, update the corresponding grid center, and repeat steps S2.2-S2.3 until the grid center no longer changes, and assign all nodes to the corresponding grid.

5. The computing power optimization method based on an industrial control gateway according to claim 1, characterized in that, The calculation process for the comprehensive performance evaluation value of each grid in step S3 is as follows: in, Indicates the first The overall performance evaluation value of each grid, Indicates the computing task in the th The completion time of each grid. This represents the shortest completion time for a computing task on the grid. This indicates the maximum completion time of a computing task on the grid. Weighting coefficients representing the completion time of computing tasks; Indicates the computing task in the th Resource utilization rate of each grid This represents the minimum resource utilization rate of a computing task on the grid. This represents the maximum resource utilization rate of computing tasks on the grid. Weighting coefficients representing the utilization rate of computing power tasks; Indicates the computing task in the th Communication latency of each grid This represents the minimum communication latency of a computing task on the grid. This represents the maximum communication latency of the computing task on the grid.

6. The computing power optimization method based on an industrial control gateway according to claim 1, characterized in that, Also includes: The real-time computing power index data of the nodes is fed back to the industrial control gateway to calculate the total comprehensive computing power score of all nodes in each grid. If the total comprehensive computing power score of each grid is above the upper limit, all nodes are assigned to the corresponding grids using the K-means algorithm.

7. A computer-readable storage medium storing a computer program, characterized in that, The computer program causes the computer to execute the computing power optimization method based on an industrial control gateway as described in any one of claims 1-6.

8. An electronic device, characterized in that, include: The system includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements the computing power optimization method based on an industrial control gateway as described in any one of claims 1-6.

Images