Research Of Production Scheduling Based On Improved Frog Leaping Algorithm

Production scheduling problem as the core of the enterprise production management and computer integrated manufacturing system, has been paid close attention to by many scholars in recent years. Making appropriate and optimized scheduling strategy can not only improve the comprehensive management level, but also bring the remarkable economic benefits. Scheduling problems have been proven to be NP hard problems. Since the traditional optimization methods can not solve the large and complicated scheduling problems, a great number of artificial intelligent methods have been introduced into the field of scheduling and great achievements have been made in recent years. Recently, with the development of computer technology and artificial intelligence technology, swarm intelligence optimization, which can generate approximate solutions close to the optimum with considerably less computational time, has become a new and effective approach to solve the production scheduling problem.This dissertation analyses typical flow shop scheduling problems, establishes the corresponding models, and proposes some swarm intelligence optimization algorithms for solving these problems. The main content of this dissertation can be summarized as follows:(1) For the blocking flow shop scheduling problem (BFSP), a new modified shuffled frog leaping algorithm (NMSFLA) is proposed to minimize the makespan. In the local steps of SFLA, a constrained crossover method is brought in for modifying the rule of frogs leaping to cope with the problem that the local search of SFLA easily generates illegal solution. The simulation results on benchmarks demonstrate the effectiveness and efficiency of the proposed NMSFLA algorithm.(2) For the flow shop scheduling problem (FSP), an extreme optimization-shuffled frog leaping algorithm (EO-SFLA) is proposed to minimize the total flow time. In EO-SFLA, the distribution rules of individual species have been detailed. In the local search process, we have simplified the jumping formula of the traditional leapfrog algorithm. At the same time, extreme optimization algorithm has been introduced. Finally, we think that each individual would keep their own jump condition of the moment before. Comparisons based on the Taillard benchmark instances indicate the superiority of the proposed EO-SFLA algorithm in terms of effectiveness and efficiency.