Multiple industrial network component rate adjustment method balancing price and task execution time factors

Abstract

With the increasingly fine and complex industrial tasks, the previous industrial tasks develop from single-layer network to multi-process network. The component rate adjustment strategy under multiple industrial network is becoming more and more complex and dynamic. In the current multiple industrial network, the traditional single network component rate adjustment strategy according to the fluctuation of electricity price cannot achieve the optimal goal of task cost of multiple network layers. Only considering the single-layer optimal component rate adjustment strategy will lead to the result of overall task timeout because of ignoring the correlation between network layers. The technical problem solved by the method is to propose a component rate scheduling method under multiple industrial networks, to consider the constraint relationship between network layers and give the component rate adjustment strategy of each network layer, and to optimize the execution cost within the task time limit.

Description

Technical Field

[0001] This invention relates to the construction of a component rate adjustment system based on multiple networks. Specifically, the system can adjust the scheduling scheme of the components executing tasks within the network layer in response to changes in the power price of the power grid, while balancing the saving of power costs with the remaining task execution time, and minimizing task costs as much as possible within the task completion time. Background Technology

[0002] With the prevalence of multi-network scenarios, component rate adjustment in industrial systems is affected by multiple coupling relationships. Existing industrial system models only consider component rate adjustment strategies for a single network layer, ignoring the interrelationships between network layers, resulting in a lack of collaborative relationships.

[0003] For the component rate adjustment problem under a single network layer in the past, an effective solution can be summarized as follows: when the electricity price is high, the component's task execution rate is greatly reduced or the component is turned off, and when the electricity price is low, the component's rate is greatly increased. Since the power consumption is reduced when the electricity price is high, this method effectively reduces the cost of the component performing the task, which has become the basic idea of ​​component rate adjustment strategy under many single industrial tasks.

[0004] In multi-network scenarios, a single network layer needs to consider not only its own task execution cost but also its relationships with other network layers. For example, if the total execution time of the overall task is fixed, excessive time spent by upper-layer networks will result in insufficient time for lower-layer networks to complete their tasks, leading to task failure. A component rate adjustment method to address this issue should consider the following facts: 1. Electricity prices fluctuate when components execute tasks; therefore, to achieve cost savings, the component operating rate within each network layer should be schedulable, and the change in component operating rate should be instantaneous without scheduling costs. 2. When a task arrives, the amount of task each network layer needs to execute should be deterministic; that is, each network layer should know the amount of task within the task timeframe. 3. Electricity price changes within the task timeframe are deterministic when the task arrives. 4. The cost of a task refers to the electricity cost incurred by the component in executing the task. 5. The task must be completed within its time limit; otherwise, the task is considered a failure. Summary of the Invention

[0005] Technical Problem: The purpose of this invention is to propose a component rate adjustment method that considers the constraints between different network layers in a multi-industry network. The scheduling objective of this method is the operating rate of components in each network layer. A fundamental constraint exists in the scheduling method: the overall task completion time. Based on this constraint, the scheduling method can complete the task within the time limit. This method solves the problem of task timeouts and failures caused by the contradiction between the component rate adjustment strategy within each network layer and the inter-layer correlation (the total task time is fixed) in a multi-industry network scenario. Furthermore, it optimizes the task execution cost to obtain an approximate globally optimal solution.

[0006] Technical Solution: In an industrial system with multiple networks, each network layer undertakes a corresponding type of industrial task. Due to fluctuations in electricity prices during task execution, the cost of component task execution changes. Furthermore, there are inter-layer coupling relationships between different network layers. Therefore, a component rate adjustment strategy considering the constraints between different network layers is introduced to rationally plan the component's operating rate. By leveraging the constraint that the total task execution time has an upper limit, the task execution time of components within each network layer is reasonably restricted. The main technical solution of this component rate adjustment method is as follows:

[0007] When a new industrial task arrives, the initial phase determines the amount of task S required to be completed by each network layer corresponding to the task S. p And the total time limit S for the task t .

[0008] When a task begins, it starts with the first network layer to be executed and proceeds through all network layers sequentially, following the top-down material flow order. Components in lower-level networks will not start running until the upper-level network has completed its task. Once the upper-level network finishes its task, its adjacent lower-level networks will immediately begin executing their assigned tasks, and so on, until the task is completed. While a network layer is executing a task, the operating speed of its components is scheduled according to the component rate adjustment strategy corresponding to that network layer until the task for that network layer is finished.

[0009] Beneficial effects:

[0010] (1) The adaptive component rate adjustment strategy for reducing the production costs of each industrial subsystem can determine the component rate adjustment strategy within the network layer by combining the constraint relationship between each network layer of the whole system, and can solve the scheduling of components within each network layer.

[0011] (2) Reduce the total cost of completing the task by dynamically adjusting the component rate adjustment strategy of each network layer of the system. The entire system formulates a scheduling strategy for each network layer based on the change of electricity price over time. This can solve the problem that the scheduling result of only considering the rate adjustment strategy of a single network component in the past is trapped in local optima, and achieve the global optimum of the total cost of the task within the effective time.

[0012] (3) Enhancing the stability and security of industrial systems In industrial systems, the excessively high operating power of some components in certain networks at a specific time due to the constraints between different network layers can lead to a decrease in the stability and security of the components, which may result in system failures due to component instability. However, the component rate adjustment strategy that takes into account the constraints between different network layers can keep the operating rate of components in each network layer within a suitable range, thereby improving the stability and security of the entire system. Attached Figure Description

[0013] Figure 1 This diagram illustrates a simplified process for natural gas production, using a natural gas production network within a circular industrial park as an example.

[0014] Figure 2 This is a schematic diagram of the main principle of the present invention. Detailed Implementation

[0015] (1) In the initial stage, the perception of multiple networks, components within network layers, inter-network layer relationships, and future electricity price changes is carried out: the multiple network perception process perceives the various attributes of each network layer and the various attributes of the industrial task; the component perception process within network layers perceives the various attributes of each component; the inter-network layer relationship perception process needs to perceive the limiting factors of material flow rate between adjacent network layers; the future electricity price change perception process needs to perceive the change in electricity price from the current time to the completion of the task. In the initial stage, we describe these four perception objects and perception processes in turn:

[0016] • Multi-network sensing process: In the initial stage, it is first necessary to analyze the specific multi-industry network scenario. The multi-industry network L consists of multiple industrial networks l p ∈N constitutes, each industrial network l p Composed of isomorphic industrial components Composition. In industrial systems with multiple industrial networks, each network layer performs different tasks. Therefore, we should also obtain basic information about the tasks. For an industrial task S, we can use S = {S...} p ,T p ,S T To describe, where S p Indicates network layer l p The amount of work required to perform task S, T p Indicates network layer lp Time spent completing the task, S T Indicates the maximum completion time of the task.

[0017] • Component awareness process within network layers: After analyzing the network structure, each component needs to be analyzed in detail. Network layer l p Any component It can be represented by a triple: in Indicates network layer l p Medium components The operating rate corresponding to electricity price level k, Indicates network layer l under electricity price level k p Medium components The runtime.

[0018] • Network Layer Relationship Awareness Process: After analyzing the components, the next step is to analyze the relationships between network layers. Since task completion time is limited in multi-network scenarios, we need to know the maximum running speed of each network layer. Based on this rate, we can find the minimum task completion time for each network layer. The specific inter-layer constraints will be given in the subsequent section on inter-layer constraints.

[0019] • The process of sensing future electricity price changes: After defining the overall network layers, inter-layer relationships, and intra-layer attributes, the next step is to sense the external variable of electricity price for the entire system. We assume that the electricity used in the multi-industry network scenario is divided into peak, off-peak, and valley electricity, and we define their prices as g. i (i = 1, 2, 3).

[0020] At the end of the initial phase, it is necessary to determine the component rate adjustment strategy within each network layer based on changes in electricity prices. The ultimate goal of the system is to optimize the cost of task execution to the greatest extent possible while successfully executing the task.

[0021] (2) Inter-layer constraints (total task time limit):

[0022] When a task begins, we need to determine the constraints between network layers. Since the total completion time of the task is fixed, the execution time of each network layer for task S has the following relationship:

[0023]

[0024]

[0025]

[0026] Among them, constraint (1) indicates that the network layer must complete the task within the time limit; constraint (2) indicates that the maximum execution time of the network layer is the total task time minus the time consumed by the upper network to complete the task and minus the minimum time required for the lower network to execute the task; constraint (3) indicates the minimum time spent by the network layer to execute the task.

[0027] (3) Optimization objective representation:

[0028] Once the constraints between network layers are determined, the necessary constraints within each network layer need to be determined, and our optimization objective—a function representing the component execution cost—can be expressed. For the duration of different electricity prices (peak, off-peak, and valley) during the network layer's runtime, we can derive...

[0029]

[0030] The network layer also needs to complete its assigned tasks, so we can derive a constraint on the amount of tasks completed by the network layer.

[0031]

[0032] Among them, constraint (4) indicates that the network layer electricity price type is divided into three types, namely peak electricity, flat electricity, and valley electricity;

[0033] Constraint (5) indicates that the network layer needs to complete the amount of tasks assigned to it.

[0034] Ultimately, we can derive the network layer l p Task cost incurred in completing the task

[0035]

[0036] The final cost of completing the entire task is

[0037]

[0038] The final solution objective is

[0039]

[0040] Let the solution set of the objective function be...

[0041]

[0042] Then the solution set Str p For network layer l p The component rate adjustment strategy indicates the network layer l p The system will adaptively adjust the component's running speed according to this strategy, and ultimately the system will run at a speed that does not exceed the task time limit, based on the strategy Str. pThe operating rates of components in each network layer are scheduled to minimize the overall task cost.

[0043] (4) Balancing electricity prices with remaining task execution time and component rate adjustment process

[0044] The process of adjusting the component rate based on electricity price: when the objective function solution result Once determined, network layer l p The component rate adjustment plan based on the electricity price can be determined from the solution results. That is, when the task starts and the electricity price is g1, the component operating rate will be set to... And set the duration of component operation at this rate to be When the electricity price changes from g1 to g2, the network layer performs component rate adjustment, changing the component's operating rate from g1 to g2. Dispatch to And set the duration of component operation at this rate to be When the electricity price changes from g2 to g3, the network layer performs component rate adjustment again, changing the component's operating rate from... Dispatch to And set the duration of component operation at this rate to be When network layer l p When the component rate adjustment plan within the layer is completed, the tasks within that layer are also completed, and the network layer is finished. p+1 The tasks within this layer will begin execution immediately, and will be carried out according to the strategy. Execution Component Network Layer p+1 The component rates within the layer are adjusted until the assigned workload for that layer is completed. This process continues until network layer l is reached. n Once all component rate adjustment plans have been executed, it means that task S has been completed and the execution cost has been minimized within the task time limit.

[0045] Since the execution time of the task may not meet the total time of the component rate adjustment strategy based on electricity price for all network layers, we should weigh the strategy against the remaining task time when determining how to adjust the component rate based on the obtained component rate adjustment strategy based on electricity price.

[0046] The component rate adjustment process that balances electricity price and remaining task time: Given the network layer l after completing the task p The actual time consumed is T p Network layer p The minimum execution time required for the task is So when If this occurs, it indicates that the remaining time of the task no longer meets the conditions for component rate adjustment based on power cost, and the remaining network layer l that has not yet executed the task...p+1 to l n Without taking electricity price changes into account and at maximum rate Complete all remaining assigned tasks, when When established, it indicates that the remaining time of the task satisfies the network layer l. p+1 Component rate adjustment conditions based on electricity cost, network layer l p+1 The component rate adjustment strategy that optimizes the power cost of this layer will continue to be implemented. p+1 Perform component rate adjustment within the layer. Network layer l p+1 After completing the task at this layer, the system needs to determine whether the remaining time of the task satisfies the requirements of the next network layer. p+2 If the conditions for component rate adjustment based on power cost are met, the component rate adjustment plan based on power cost will continue to be executed; otherwise, the remaining network layers that have not yet executed tasks will execute all remaining tasks at the maximum rate, and so on, until the task is completed.

Claims

1. A component rate adjustment method considering inter-layer constraints in a multi-layer industrial network, characterized in that: In the initial stage, a specific multi-industry network scenario is analyzed. Composed of multiple industrial network layers Composition, each industrial network Composed of isomorphic industrial components Composition, for industrial tasks use To depict, among which Represents network layer Corresponding task The amount of tasks to be performed Represents network layer Time spent completing the task Indicates the maximum completion time of the task; After analyzing the network structure, each component is analyzed in detail, including the network layers. Any component Represented by a triple: in Represents network layer Medium components Corresponding electricity price level The running speed, Indicates electricity price level Lower network layer Medium components The runtime; After a detailed analysis of the components, the relationships between the network layers were analyzed to determine the maximum operating speed of each network layer. Based on this rate, the minimum task completion time to each network layer. ; After defining the overall network layers, inter-layer relationships, and intra-layer attributes, the external variable of electricity price for the entire system is perceived. It is assumed that the electricity used in the multi-industry network scenario is divided into peak electricity, off-peak electricity, and valley electricity, and their prices are defined as follows: After defining the overall network layers, inter-layer relationships, and intra-layer attributes, the external variable of the entire system, electricity price, is perceived, assuming the electricity energy used in a multi-industry network scenario. When the task begins, for the task S The task execution times of each network layer are related as follows: Among them, constraint (1) indicates that the network layer must complete the task within the time limit; constraint (2) indicates that the maximum execution time of the network layer is the total task time minus the time consumed by the upper network to complete the task and minus the minimum time required for the lower network to execute the task; constraint (3) indicates the minimum time spent by the network layer to execute the task.

2. The component rate adjustment method considering inter-layer constraints in a multi-layer industrial network according to claim 1, characterized in that: Once the constraints between network layers are determined, the necessary constraints within each network layer need to be determined, and the optimization objective—a function of component execution cost—needs to be expressed. This involves determining the duration of different electricity prices during the network layer's runtime. The network layer also needs to complete its assigned tasks, thus deriving a constraint on the amount of tasks completed by the network layer. Among them, constraint (4) indicates that the network layer electricity price type is divided into three types, namely peak electricity, flat electricity, and valley electricity; Constraint (5) indicates that the network layer needs to complete the amount of tasks assigned to it. Finally, the network layers are derived. Task cost incurred in completing the task The final cost of completing the entire task is The final solution objective is Let the solution set of the objective function be... Then the solution set For network layer The component rate adjustment strategy indicates the network layer The system will adaptively adjust the component's running speed according to this strategy, and ultimately the system will operate within the time limit set by the strategy. The operating rates of components in each network layer are scheduled to minimize the overall task cost.

3. The component rate adjustment method considering inter-layer constraints in a multi-layer industrial network according to claim 2, characterized in that: When the solution set of the objective function After it is generated, the network layer The component rate adjustment plan within the system can be determined based on the solution results, i.e., when the electricity price is... At that time, the component's running speed will be set to And set the duration of component operation at this rate to be When the electricity price is Change to At that time, the network layer executes a component rate adjustment plan, reducing the component's operating rate from... Dispatch to And set the duration of component operation at this rate to be When the electricity price is Change to At that time, the network layer continues to execute the component rate adjustment plan, reducing the component's operating rate from... Dispatch to And set the duration of component operation at this rate to be When the network layer When the component rate adjustment plan within the network layer is completed, the tasks within that layer are also completed. The tasks within this layer will begin execution immediately, and will be carried out according to the strategy. Execution Component Network Layer The component rate is adjusted until the task assigned to this layer is completed, and so on, until the network layer. Once all component rate adjustment plans have been executed, the task is considered complete. The task was completed successfully, and the execution cost was minimized within the task timeframe. The network layer that has completed the task is known. The actual time consumed is Network layer The minimum execution time required for the task is , then when If this occurs, it indicates that the remaining time of the task no longer meets the conditions for component rate adjustment based on power cost, and the remaining network layer tasks that have not yet been executed... arrive Without taking electricity price changes into account and at maximum rate Complete all remaining assigned tasks, when Upon establishment, it indicates that the remaining task time meets the condition for component rate adjustment based on power cost, network layer. The component rate adjustment strategy that optimizes the power cost of this layer will continue to be implemented. To perform component rate adjustment within a layer, after each network layer completes its task, the system must determine whether the remaining time of the task meets the conditions for component rate adjustment based on power cost in the next network layer. If it does, the component rate adjustment plan based on power cost will continue to be executed; otherwise, all remaining tasks will be completed at the maximum rate until the task ends. The component rate adjustment process that balances electricity price and remaining task time: Given the network layer that has completed its task. The actual time consumed is Network layer The minimum execution time required for the task is , then when If this occurs, it indicates that the remaining time of the task no longer meets the conditions for component rate adjustment based on power cost, and the remaining network layer tasks that have not yet been executed... arrive Without taking electricity price changes into account and at maximum rate Complete all remaining assigned tasks, when When established, it indicates that the remaining time of the task satisfies the network layer requirements. Component rate adjustment conditions based on electricity cost, network layer The component rate adjustment strategy that optimizes the power cost of this layer will continue to be implemented. Adjust the component rate within the layer, network layer After completing the task at this layer, the system needs to determine whether the remaining time for the task satisfies the requirements of the next network layer. If the conditions for component rate adjustment based on power cost are met, the component rate adjustment plan based on power cost will continue to be executed; otherwise, the remaining network layers that have not yet executed tasks will execute all remaining tasks at the maximum rate, and so on, until the task is completed.

Images