Application And Research On Knowledge Database With Learning Mechanism Of Genetic Algorithm On Job-Shop Scheduling

In recent years, rapid economic development in the manufacturing industry and more intense competition in the market. How to maintain the quality of products under the conditions with high efficiency, low cost production tasks to accomplish this goal, becoming the focus of manufacturers market share. This requires that manufacturers be able to make reasonable arrangements shop scheduling processes, rational use of resources to reduce the duration and reduce production costs. Therefore, the shop scheduling problem is receiving increasing attention.Shop scheduling problem belongs to NP-hard problem, is a typical optimization problem solving the most difficult problems. Genetic algorithms for its versatility, the algorithm and simple features, has been widely applied to the optimization of shop scheduling problems. However, it is because of its versatility, leading to poor flexibility. As the algorithm is simple, although can guarantee global convergence, but can not be avoided will be a partial degradation. Constrained genetic algorithm to solve job shop scheduling application in the bottleneck problem is many of today's experts and scholars on the main topics.This paper presents a critical path based on cross-compression mechanism, using the mechanism, can effectively increase the cross-operation is to obtain good probability of an individual so that the whole evolution can be faster to have the approximate optimal solution. This paper, the learning mechanism introduced into the genetic algorithm, the use of machine learning memory and storage capabilities of a knowledge base to optimize the process and to optimize the results of classification is stored in order to guide the future of the shop scheduling, making retrieval more efficient, thereby establishing a high average fitness value of the initial population. At the same time, using real-time learning mechanism can also guide the genetic algorithm crossover and mutation parameters, thus effectively making the evolutionary process along on better direction.This paper designs and implements a shop scheduling knowledge base system platform. In-depth study and knowledge of the job shop scheduling theory, optimization models, the improved algorithm will be presented by Muth and Thompson benchmarks tests, results show that the optimization algorithm has higher efficiency and to understand the quality and convergence speed. And apply the improved algorithm model for solving practical problems, the results obtained is feasible and effective.