Design Of Parallel Petri Nets For Supervising And Scheduling Pipeline Systems

Pipeline systems are widely used to carry out production and transportation tasks.In major national projects such as chemical industry,metallurgy and energy,large pipeline systems are required to connect the main equipment with supporting equipment,and control the fluid transport status in the pipeline system through various control valves.Therefore,the overall performance of the pipeline system will directly affect the operating efficiency of the entire industrial process control system.How to ensure the safe and efficient operation of the pipeline system is very necessary.However,there are a large number of control valves and transmission pipelines in the pipeline system,which brings great challenges to our research on optimal dispatch and control.Parallel system is a method proposed in recent years to study complex system problems,and has been applied in some fields.Based on the idea of parallel system,a design method of parallel Petri net for pipeline control and dispatch is presented in this paper.The main research work of this paper is as follows:1.Establishment of parallel Petri net for pipeline system.The graphical and distributed nature of Petri nets makes it a good tool for modeling and integrating discrete event systems.In this paper,the artificial pipeline system is built using Petri net,and the parts are modeled step by step.The logical relationship between production operations in the pipeline system is analyzed,and the production process is modeled modularily according to the proposed concept of logical operation.Considering the conflict relationship between production operations,use case diagrams are used as a graphical method to find out the conflict set,which transforms the conflict set into general linear inequality constraints,and adds a constraint structure to the model.The mapping relationship of the actual operation makes it possible to control the actual pipeline system in parallel.2.On the basis of the parallel Petri net of the pipeline system,the scheduling strategy is studied through computational experiments.This paper presents a node information calculation method for reachable and reverse trees of parallel Petri nets to obtain the evolving data of the system's running state.A heuristic function h(m)in the A* algorithm is fitted using the derived evolutionary dataset through a deep neural network.The function is used to estimate the time cost from each running state to the target state of the system,and then the A* algorithm is used to solve a reasonable and efficient scheduling strategy.3.The CIF3 tool is used to build a two-dimensional visualization of a parallel pipeline simulation system that evolves and executes in parallel.The automaton models of the actual pipeline system and the parallel Petri nets and their SVG vector diagrams are established respectively,and the scheduling strategy obtained from the calculation experiment is applied to the parallel pipeline simulation system for evolution experiments.Through analysis and demonstration experiments,the feasibility of the proposed method is verified.