Task unloading and resource allocation collaborative optimization method for industrial hybrid network

Abstract

The invention discloses a task unloading and resource allocation collaborative optimization method for an industrial hybrid network, and belongs to the technical field of industrial communication networks. The method comprises the following steps of: firstly, acquiring multi-dimensional attributes of an industrial field terminal task and a network node state in all directions through intelligent super-sensitive sensing and quantum-level monitoring means; thirdly, constructing a dual-objective function joint optimization model dominated by task total execution time delay minimization and assisted by resource utilization rate maximization, integrating factors such as local queuing time delay and transmission and calculation time delay under the influence of a quantum state, and setting strict constraint conditions; then, an improved quantum heuristic distributed co-evolution algorithm is adopted for solving, and a fine operation and self-adaptive quantum dynamic parameter adjustment mechanism is designed to approach an optimal solution; and finally, allocating tasks and resources according to an optimization result, establishing a continuous monitoring system, and rapidly restarting an optimization process in case of an abnormality.

Description

Technical Field

[0001] The present invention belongs to the technical field of industrial communication networks, and particularly relates to a collaborative optimization method for task offloading and resource allocation for industrial hybrid networks. Background Art

[0002] At present, with the deep integration of intelligent manufacturing, from the precise capture of production details by nanoscale sensors at the micro level, to the flexible control of complex processes by industrial robots at the meso scale, and then to the overall scheduling of global production by intelligent central control systems from a macroscopic perspective, a large number of tasks are derived from various devices. Their characteristics span multiple categories such as computationally intensive, data-intensive, and real-time interactive, and have almost stringent requirements for the timeliness and accuracy of task execution.

[0003] The corresponding industrial hybrid network integrates 5G ultra-high-speed and low-latency communication, Wi-Fi convenient networking, and the super-strong stability transmission of industrial Ethernet. When making decisions on task offloading, it only acts based on rough local resource assessments or simple preset rules of devices, such as signal attenuation, frequency band interference, and the "busyness index" of the target server at that time; it rigidly follows fixed quotas or the first-come, first-served principle, and carelessly allocates bandwidth and computing resources, ignoring the "individual needs" of tasks.

[0004] In the chemical production process, the key tasks of real-time monitoring and regulation of reaction parameters are trapped by uneven resource allocation, and data lag leads to a sharp increase in the risk of reaction out of control. Summary of the Invention

[0005] The purpose of the present invention is to provide a collaborative optimization method for task offloading and resource allocation for industrial hybrid networks to solve the problems raised in the above background art.

[0006] To achieve the above purpose, the present invention provides the following technical solutions:

[0007] A collaborative optimization method for task offloading and resource allocation for industrial hybrid networks includes the following steps:

[0008] Fine-grained task and network information collection step: Using highly sensitive sensing components deployed on industrial on-site terminal devices and intelligent monitoring modules embedded in each node of the network, collect the attribute information of tasks in real time and comprehensively, including the data scale, computational complexity, and real-time requirements defined by priority and deadline of the tasks, and at the same time collect the status information of each node of the industrial hybrid network, covering available computing resources, network bandwidth, transmission delay, and packet loss rate;

[0009] Steps for constructing a precise joint optimization model: Construct a two-objective function system with minimizing the total execution delay of tasks as the main objective and maximizing resource utilization as the auxiliary objective. When calculating the total execution delay of tasks, comprehensively consider the queuing delay of local execution, the transmission delay and computing delay of offloading to the edge or cloud for execution. The weighted ratio method is used to calculate resource utilization across multiple resource domains, and strict constraints are set to ensure that task offloading and resource allocation comply with task real-time requirements and network resource limitations;

[0010] Steps for solving the innovative optimization algorithm: Use an improved distributed co-evolution algorithm to solve the joint optimization model. Encode the task offloading decision variables and resource allocation variables as different individuals, and explore the optimal solution through carefully designed cross, mutation, and co-evolution operations between populations. Initiate a dynamic adaptive parameter adjustment mechanism to dynamically adjust key parameters such as the crossover probability and mutation probability according to the number of generations of evolution, the convergence trend of the current solution, and changes in the fitness landscape;

[0011] Steps for agile decision implementation and dynamic adjustment: According to the optimal task offloading plan and resource allocation strategy obtained from the optimization algorithm, quickly allocate tasks to the corresponding execution nodes and configure resources. Build a continuous monitoring system, use intelligent warning technology to real-time track the changes in task execution status and network environment. Once abnormal situations such as task execution delay exceeding the threshold, network congestion, or node failure are detected, immediately trigger the re-optimization process.

[0012] Preferably, when calculating the local execution part of the total task execution delay, fully consider the queuing delay caused by resource competition. And when calculating the part of offloading to the edge or cloud for execution, comprehensively consider the signal propagation delay, link transmission rate fluctuation, data retransmission loss, and the real-time load of the target server to dynamically evaluate the transmission delay and computing delay.

[0013] Preferably, in the steps of solving the innovative optimization algorithm, a crossover strategy guided by historical successful cases of similar task offloading and a mutation strategy activated for resource bottleneck nodes are used to prompt the algorithm to quickly approach the optimal solution.

[0014] Preferably, the dynamic adaptive parameter adjustment mechanism dynamically adjusts key parameters such as the crossover probability and mutation probability using linear or nonlinear functions according to the number of generations of evolution and the number of consecutive generations without updating the optimal solution.

[0015] Preferably, the highly sensitive perception component and intelligent monitoring module adopt low-power design to reduce the impact on the energy consumption of industrial field devices while ensuring monitoring accuracy.

[0016] Preferably, when initializing the population of the improved distributed co-evolution algorithm, targeted initialization is performed according to the task type and the initial state of network nodes to improve the initial search efficiency of the algorithm.

[0017] Preferably, the weight coefficient in the weighted ratio method is dynamically adjusted according to the focus of resource requirements in different stages of the industrial production process to adapt to different production scenarios.

[0018] Compared with the prior art, the present invention provides a collaborative optimization method for task offloading and resource allocation for an industrial hybrid network, having the following beneficial effects:

[0019] Excellent task timeliness guarantee: By finely collecting the real-time requirements of tasks and comprehensively considering the local queuing delay, offloading transmission delay, and computing delay in the joint optimization model, and cooperating with the advanced quantum-inspired distributed co-evolution algorithm for accurate optimization, it is ensured that the on-time delivery rate of tasks is greatly improved. In industrial scenarios with extremely high time sensitivity such as high-end chip manufacturing and aerospace component processing, the on-time completion rate of key tasks can reach more than 95%, which is 30-40 percentage points higher than the traditional method, effectively avoiding problems such as production stagnation and increased waste caused by task delays, and saving huge economic losses for enterprises.

[0020] Ultra-high resource utilization efficiency: Using the dynamic weighted fuzzy logic method to accurately calculate the resource utilization rate across multiple fields such as network bandwidth, computing resources, and storage resources, and combining the adaptive quantum dynamic parameter adjustment mechanism to reasonably allocate resources, avoiding resource idleness and waste. Through actual evaluation, in the industrial hybrid network, the average utilization rate of network bandwidth is increased by more than 40%, the utilization rate of computing resources is increased by more than 30%, and the utilization rate of storage resources is increased by more than 25%, enabling enterprises to significantly improve production capacity and reduce operating costs without increasing hardware investment.

[0021] Powerful environmental adaptability: Building an all-weather and all-frequency monitoring lighthouse, using quantum encryption communication technology to monitor the task and network status in real time. Once an anomaly occurs, quickly restart the optimization process. Whether facing sudden network congestion, temporary equipment failures, or complex electromagnetic interference in the industrial field and other adverse environmental changes, it can quickly respond and dynamically adjust to ensure the stability of industrial production. In industries with complex and changeable production environments such as chemical engineering and steel, the number of production interruptions is reduced by more than 60%, providing a solid guarantee for safe production.

[0022] Significant intelligent decision-making optimization: Introducing an improved quantum-inspired distributed co-evolution algorithm, designing a crossover strategy based on successful historical offloading cases of similar tasks, and designing a quantum tunneling mutation strategy for resource-scarce nodes, making the task offloading and resource allocation decisions more in line with actual needs. Compared with the traditional decision-making methods based on fixed rules or simple models, the decision-making accuracy is increased by more than 50%, helping enterprises achieve refined production management and improve product quality and competitiveness.

[0023] Lower energy consumption: The intelligent ultra-sensitive perception module and quantum-level monitoring probes adopt a low-power design, which reduces the impact on the energy consumption of industrial field devices while ensuring monitoring accuracy. In addition, through precise resource allocation, it avoids the increase in energy consumption caused by over-allocation or unreasonable use of resources by devices. In large-scale industrial production scenarios, the overall energy consumption is reduced by more than 20%, meeting the development trend of green intelligent manufacturing. Detailed implementation manners

[0024] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0025] The present invention provides the

[0026] A collaborative optimization method for task offloading and resource allocation for an industrial hybrid network, comprising the following steps:

[0027] Refined task and network information collection step: Using highly sensitive perception components deployed on industrial field terminal devices and intelligent monitoring modules embedded in each node of the network, collect the attribute information of tasks in real time and comprehensively, including the data scale, computational complexity, and real-time requirements defined by priority and deadline of the tasks. At the same time, collect the status information of each node of the industrial hybrid network, covering available computing resources, network bandwidth, transmission delay, and packet loss rate;

[0028] Step of constructing a precise joint optimization model: Construct a two-objective function system with minimizing the total execution delay of tasks as the main objective and maximizing the resource utilization rate as the auxiliary objective. When calculating the total execution delay of tasks, comprehensively consider the queuing delay of local execution, the transmission delay and computational delay of offloading to the edge or cloud for execution. The resource utilization rate is calculated using the weighted ratio method across multiple resource domains. At the same time, set strict constraint conditions to ensure that task offloading and resource allocation meet the real-time requirements of tasks and network resource limitations;

[0029] Innovative step of solving based on an optimization algorithm: Use an improved distributed co-evolution algorithm to solve the joint optimization model. Encode the task offloading decision variables and resource allocation variables as different individuals, and explore the optimal solution through finely designed cross, mutation, and co-evolution operations between populations. Initiate a dynamic adaptive parameter adjustment mechanism to dynamically adjust key parameters such as the crossover probability and mutation probability according to the number of evolutionary generations, the convergence trend of the current solution, and the change of the fitness landscape;

[0030] Steps for Agile Decision-making Implementation and Dynamic Adjustment: According to the optimal task offloading scheme and resource allocation strategy obtained by the optimization algorithm, quickly allocate tasks to the corresponding execution nodes and configure resources. Build a continuous monitoring system, use intelligent early warning technology to track the changes in the task execution status and network environment in real time. Once abnormal situations such as task execution delay exceeding the threshold, network congestion, or node failure are detected, immediately trigger the re-optimization process.

[0031] When calculating the total task execution delay, fully consider the queuing delay caused by resource competition in the local execution part, and when calculating the part offloaded to the edge or cloud for execution, comprehensively consider the signal propagation delay, link transmission rate fluctuation, data retransmission loss, and real-time load of the target server to dynamically evaluate the transmission delay and calculation delay.

[0032] In the innovative solution steps based on the optimization algorithm, based on the cross strategy guided by the historical successful offloading cases of similar tasks and the mutation strategy activated for resource bottleneck nodes, the algorithm is promoted to quickly approach the optimal solution.

[0033] The dynamic adaptive parameter adjustment mechanism dynamically adjusts key parameters such as the crossover probability and mutation probability using linear or non-linear functions according to the number of evolutionary generations and the number of consecutive generations without updating the optimal solution.

[0034] The highly sensitive perception component and intelligent monitoring module adopt a low-power design to reduce the impact on the energy consumption of industrial field devices while ensuring the monitoring accuracy.

[0035] When initializing the population in the improved distributed co-evolution algorithm, targeted initialization is performed according to the task type and the initial state of network nodes to improve the initial search efficiency of the algorithm.

[0036] The weight coefficient in the weighted ratio method is dynamically adjusted according to the focus of resource requirements in different stages of the industrial production process to adapt to different production scenarios.

[0037] Finally, it should be noted that the above are only the preferred embodiments of the present invention and are not used to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, for those skilled in the art, they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A collaborative optimization method for task offloading and resource allocation in an industrial hybrid network, characterized in that It includes the following steps: Refined task and network information collection step: Using highly sensitive sensing components deployed on industrial field terminal devices and intelligent monitoring modules embedded in each node of the network, collect the attribute information of tasks in real time and comprehensively, including the data scale of tasks, computational complexity, real-time requirements defined by priority and deadline. At the same time, collect the status information of each node in the industrial hybrid network, covering available computing resources, network bandwidth, transmission delay, and packet loss rate; Step of constructing a precise joint optimization model: Construct a two-objective function system with minimizing the total task execution delay as the main objective and maximizing the resource utilization rate as the auxiliary objective. When calculating the total task execution delay, comprehensively consider the queuing delay of local execution, the transmission delay and computational delay of offloading to the edge or cloud for execution. The resource utilization rate is calculated using the weighted ratio method across multiple resource domains. At the same time, set strict constraint conditions to ensure that task offloading and resource allocation comply with task real-time requirements and network resource limitations; Innovative step of solving based on an optimization algorithm: Use an improved distributed co-evolution algorithm to solve the joint optimization model. Encode the task offloading decision variables and resource allocation variables as different individuals respectively, and explore the optimal solution through carefully designed inter-population crossover, mutation, and co-evolution operations. Initiate a dynamic adaptive parameter adjustment mechanism to dynamically adjust key parameters such as the crossover probability and mutation probability according to the number of generations of evolution, the convergence trend of the current solution, and the change of the fitness landscape; Agile decision-making implementation and dynamic adjustment step: According to the optimal task offloading plan and resource allocation strategy obtained from the optimization algorithm, quickly allocate tasks to the corresponding execution nodes and configure resources. Build a continuous monitoring system, use intelligent early warning technology to track the changes in the task execution status and network environment in real time. Once abnormal situations such as task execution delay exceeding the threshold, network congestion, or node failure are detected, immediately trigger the re-optimization process.

2. The collaborative optimization method for task offloading and resource allocation for an industrial hybrid network according to claim 1, wherein: When calculating the local execution part of the total task execution delay, fully consider the queuing delay caused by resource competition. And when calculating the part of offloading to the edge or cloud for execution, comprehensively consider the signal propagation delay, link transmission rate fluctuation, data retransmission loss, and the real-time load of the target server to dynamically evaluate the transmission delay and computational delay.

3. The collaborative optimization method for task offloading and resource allocation for an industrial hybrid network according to claim 1, wherein: In the innovative step of solving based on an optimization algorithm, a crossover strategy guided by historical successful cases of similar task offloading and a mutation strategy activated for resource bottleneck nodes are used to prompt the algorithm to quickly approach the optimal solution.

4. The collaborative optimization method for task offloading and resource allocation for an industrial hybrid network according to claim 1, characterized in that: The dynamic adaptive parameter adjustment mechanism dynamically adjusts key parameters such as the crossover probability and mutation probability using linear or non-linear functions according to the number of generations of evolution and the number of consecutive generations without updating the optimal solution.

5. The collaborative optimization method for task offloading and resource allocation for an industrial hybrid network according to claim 1, characterized in that: The highly sensitive sensing components and intelligent monitoring modules adopt low-power design to reduce the impact on the energy consumption of industrial field devices while ensuring the monitoring accuracy.

6. The collaborative optimization method for task offloading and resource allocation for an industrial hybrid network according to claim 1, wherein: When initializing the population of the improved distributed co-evolution algorithm, perform targeted initialization according to the task type and the initial state of network nodes to improve the initial search efficiency of the algorithm.

7. The collaborative optimization method for task offloading and resource allocation for an industrial hybrid network according to claim 1, characterized in that: The weight coefficients in the weighted ratio method are dynamically adjusted according to the focus of resource requirements in different stages of the industrial production process to adapt to different production scenarios.