| With the development of intelligent manufacturing,the transformation and upgrading of traditional manufacturing industry is facing great challenges.The reasonable allocation of resources and energy is crucial to the upgrading and development of manufacturing industry.In the actual production system,the production scheduling plan and equipment maintenance plan interact with each other.The production process will make the equipment aging and increase the failure rate of the equipment.It is necessary to carry out preventive maintenance on the equipment to reduce the failure rate and ensure the stable operation of the equipment.But the maintenance operation is that the equipment is not available,which leads to the extension of the production cycle and affects the deadline or product delay.Therefore,coordinating and balancing the production scheduling plan and equipment maintenance plan can improve the efficiency of the manufacturing system,protect the equipment effectively,reduce the production cost,and realize the optimal utilization of resources and energy.This thesis focuses on the batch scheduling and equipment maintenance of parallel units:(1)By studying the knowledge of parallel machine batch scheduling and preventive maintenance theory,the reliability theory and preventive maintenance knowledge of equipment are introduced.Considering that the failure function of the equipment obeys Weibull distribution,using the service life and failure rate of the equipment,the joint decision-making of multi-order and multi-type jobs and preventive maintenance for parallel equipment is made.To sum up,in order to minimize the sum of tardiness penalty and preventive maintenance cost,a mathematical model of parallel machine batch scheduling and preventive maintenance is established based on the unavailability constraints of equipment maintenance and the relationship between maintenance start time and equipment service life.(2)Based on the basic flow of genetic algorithm,an improved genetic algorithm with predatory search strategy and standard batch rule is designed.The chromosome showing the sequence of batch processing position and batch processing assignment is designed.The selection operator adaptive to population characteristics,crossover operator and mutation operator based on standard batch are designed.The selection operator with adaptive factor can realize the global search in the early stage of iteration and increase the selection pressure in the late stage of iteration,increase the proportion of high-quality solutions and improve the search accuracy of the algorithm;The crossover operator and mutation operator based on standard batch can effectively increase the population diversity;The predatory search strategy can balance the global search and local search,and realize the combination of search breadth and depth.(3)According to the analysis of experimental results,several different sizes of orders are set to verify the accuracy of the model and the efficiency of the algorithm.Through the method of standard batch,the binding of batch and external order time information and the fusion of internal algorithm are realized.In order to test the superiority and effectiveness of GA based on standard batch rule compared with the previous algorithms to solve this kind of problem,the target value convergence experiment based on iteration and the experiment of algorithm running time are designed.The results show that GA based on standard batch rule performs better in target convergence and algorithm running time,and it also can effectively reduce tardiness penalty,maintenance cost and resource waste in production scheduling and maintenance process. |
