New Model And Optimization Algorithm For Position-Based Group And Batch Scheduling

With the in-depth development of economic globalization,informatization and the rapid take-off of Chinese economy,competition among enterprises continues to intensify,and customer needs are diversified.This ever-changing market environment forces enterprises to coordinate their own resources and improve their production efficiency in order to improve their enterprises’ market competitiveness.Production scheduling directly controls the production process of an enterprise and affects the ultimate benefit of the enterprise.Facing the current market challenges,how to quickly organize enterprises to produce,quickly respond to the widespread needs of consumers,and use modeling and optimization algorithms to improve the new production scheduling technology of enterprises,so as to improve the overall production level of Chinese manufacturing industry,it has brought unprecedented new challenges to the manufacturing industry and academia.This paper first introduces the research background and significance of scheduling,sorts out some domestic and foreign research statuses related to this topic,and puts forward the basic group and batch scheduling problem to be studied.Secondly,in view of the actual processing situation in the enterprise production,this paper proposes two generalized location-based De Jong’s learning effect and Pegel’s learning effect of group and batch scheduling new models.Both models can be expressed as a basic processing model and a learning processing model with different values of the M factor in the function.When expressed as a learning situation,as the processing position of the workpiece continues to move backward,the actual processing time of the later processing workpiece will not tend to zero.Compared with the traditional scheduling model that only studies the learning effect,these two models are more in line with the actual production scheduling process,which is also an important innovation of this paper.In addition,considering the fact that enterprises still need to spend a certain amount of time to prepare various production resources to better process orders before processing,this paper also introduces production preparation time in the general scheduling model,forming two new types of scheduling model,in order to better fit the actual production conditions.Then,considering the different batch processing modes,this paper proposes four types of models based on De Jong’s and Pegel’s learning for workpieces in serial batch processing and De Jong’s and Pegel’s learning for workpieces in parallel batch processing,analyzes the standalone and parallel machine environments eight types of scheduling models are closer to the actual production of the enterprise.For the single-machine environment,through analysis and solution,the optimal algorithms for four types of problems are given;for the parallel machine environment,after problem analysis,heuristic algorithms are designed.Finally,this paper gives the detailed calculation process of the optimal algorithm in the single machine environment and tests a large amount of data to evaluate the performance of the heuristic algorithm in the parallel machine environment.From a practical perspective,the effectiveness and feasibility of the algorithm are also verified.