Research Of Improved Hopfield Neural Network Algorithm On Single Dynamic Scheduling Problem

Job-Shop scheduling problem, as the core of manufacturing business, is a class of time constraints, sequence constraints, and resource constraints of the combinatorial optimization problem. Making reasonable schedule to tasks in workshop is useful for achieving the rational allocation of corporate resources, improving labor productivity and utilization of processing equipment, reducing the production costs and so on, thus Job-shop scheduling research is very important, and has been one of the most of the gravity research field in the present. In recent years, most Job-shop scheduling research aimed at the scheduling with complexity, and the research methods is various, but the use of the improved Hopfield neural network algorithm is a new way to solve the single machine dynamic scheduling problem. In this paper, through analyzing the various constraints, establishing the optimization model of scheduling problem, and in order to achieve the goal of make the total weighted time of completion is minimum, this research used an improved Hopfield neural network algorithm made a dynamic scheduling. The main content in this article is as follows:(1) The research of dynamic Job-shop scheduling:By comparing and analyzing the current research situation, differences, respective advantages and disadvantages of the Job-shop scheduling problems, this article mainly made an outline of the current research situation and shortcomings of the general and single-machine Job-shop scheduling problems, described the classification of Job-shop scheduling and the scheduling model as well as made a brief description of this research content.(2) The theory research of Job-shop scheduling method:Made an overview of the theoretical knowledge of Job-shop scheduling method as well as the most popular heuristic scheduling algorithms, such as artificial neural network algorithm, simulated annealing algorithm and genetic algorithm which were used in this paper, and analyzed the merits and inadequate of various algorithms, in light of the characteristics of Job-shop scheduling, designed the improved Hopfield neural network algorithm which combined with the discrete Hopfield neural network and the simulated annealing algorithms.(3) Scheduling implementation methods and system research:According to the dynamic single machine scheduling problem described in this article, combined with the scheduling constraints and the proposed algorithm, established the corresponding mathematical model, and described the concrete steps for solving the issues. Finally, base on the theory above, this article developed a workshop scheduling system, achieved the Job-shop scheduling process visualization, compared and analyzed the final results of the scheduling algorithms.In the end, summarized the main results and shortcomings of this research, analyzed and forecasted the next study, and presented my own views.