Research On Resource Allocation Optimization Of Manufacturing Systems Based On Timed Petri Nets

In recent years,"Made in China 2025" and the "14th Five-Year Plan" have been put forward to accelerate the pace of building a strong manufacturing country and developing advanced manufacturing industries.However,the manufacturing industry in China still has problems,such as low resource utilization efficiency and low productivity.In general,it is still at a large but not strong stage.Various resources in manufacturing systems realize the functions of processing,transportation,loading,etc.Reasonable selection or allocation of resources can effectively improve the productivity of systems.Therefore,it is of great scientific significance and practical value to study the optimal allocation of buffer capacity and machine quantity and type in manufacturing systems with limited cost constraints to maximize system productivity.However,almost all existing studies on optimal resource allocation require traversing all states of the system,and the allocation efficiency is low.This paper takes manufacturing systems(e.g.,series-parallel production lines,assembly lines,etc.)as the research background and investigates the buffer and server quantity and type allocation problem based on timed Petri nets.The specific contents are as follows:(1)Research on the modeling and performance evaluation methods of manufacturing systemsAccording to the resource composition and workflow of manufacturing systems,the place of Timed Petri nets is divided into different subsets,and a generalized manufacturing system modeling method is proposed to construct Timed Petri nets of single batch processing and multi batch processing manufacturing systems.Then,the performance evaluation methods for these two types of manufacturing systems are outlined based on the established Timed Petri nets,which lay the foundation for studies of resource allocation problems.(2)Research on the allocation problem of manufacturing systems based on heuristic methodsCombining with the performance evaluation method,a mixed integer linear programming problem is developed to determine the optimal allocation scheme of buffer capacities and machine quantities and the corresponding productivity under a given machine type allocation.Then,a heuristic resource optimization allocation method based on the bottleneck loop analysis is developed with productivity maximization as the objective function,which can reasonably allocate buffer capacity and the quantity and type of machines to maximize system productivity.And the design and implementation of the heuristic resource allocation method are shown concretely by two simulation examples.(3)Research on the allocation problem of manufacturing systems with batch productions based on hybrid genetic algorithmsManufacturing systems with batch productions,where each process can process single or multiple similar products at a time,are widely used but also have more challenging resource allocation problems.To address this problem,the rationality of chromosomes is described by combining the activity determination conditions of Petri nets and the specific constraints of the problem,and a hybrid genetic algorithm is designed by combining a genetic algorithm with a simulated annealing algorithm with the objective function of maximizing productivity,which can reasonably allocate the buffer capacity and the quantity and type of machines to achieve the maximization of the productivity of systems.The superiority of the hybrid genetic algorithm compared with the genetic algorithm in terms of allocation results is verified by simulation experiments of the manufacturing system.(4)Research on the allocation problem of manufacturing systems with batch productions based on improved hybrid genetic algorithmsTo address the problem that the optimal resource allocation results of large manufacturing systems obtained by hybrid genetic algorithms are not satisfactory due to the large solution space,a mixed integer linear programming problem is developed to determine the optimal allocation scheme of machine types and corresponding productivity under a given buffer capacity and machine quantity allocation.Combining it with the hybrid genetic algorithm to obtain the improved hybrid genetic algorithm,which can greatly reduce the solution space and further improve the resource allocation scheme.The simulation results show that the improved hybrid genetic algorithm proposed in this paper improves the solution quality by 9.89%compared to the hybrid genetic algorithm.In summary,this paper proposes the corresponding joint resource allocation method of buffer capacity and quantity and type of machines for manufacturing systems without and with batch processing that can efficiently and reasonably allocate buffer capacity,quantity,and type of machines to maximize system performance.It provides effective decision support in the stages of planning new plants and continuous improvement of resource allocation for manufacturing companies.