Solving Robust Flow-Shop Scheduling Problems With Uncertain Processing Times Based On Hybrid Particle Swarm Optimization Algorithm

With the development of global economic integration, competition has become more and more fiercely among the business enterprises. To program operation way for production reasonably and economize production cost are the keys for a company to survive and grow larger. However, a best scheduling is very important to program operation way for production reasonably. Therefore, the research on product scheduling has not only tremendous academic value, but also has great practical meaning.Flow-shop Scheduling (FSSP) is a kind of production scheduling. In the real production process, much uncertainty exists due to various factors.This paper mainly research on FSSP with uncertain processing times, considering both uncertainty and reducing the affect of the disturbance as much as possible. Particle Swarm Optimization (PSO) is a newly intelligent optimization algorithm, and it has notable characteristics, such as simple principle, less parameters and operations, moreover, it is easily realized. So it is a kind of efficient parallel search algorithm. Presently, PSO has been widely used in many fields, for example, function optimization, manual neural network training, combination optimization, etc. and it has been a new focus of evolution algorithms research. The domain of research on PSO is larger and deeper. Based on the research on PSO and other intelligent optimization algorithms, a hybrid PSO was proposed, which was used to deal with FSSP with uncertain processing times. The main tasks of the paper are as follows:(1)The principles and characteristics of PSO were analyzed, including parameters in PSO. Improved methods and applications of the current PSO were summarized. Based on the characteristics of PSO, an improved PSO algorithm was proposed which used the dominance of other algorithms to remedy the shortcomings of the PSO.(2) A robust makespan criterion was proposed based on the FSSP with uncertain processing times. Through compromising the two conflict objectives, a more robust scheduling with minimizing makespan is obtained. Based on therobust makespan criterion, extensive experiments were performed onsingle-objective and bi-objective problems using the hybrid PSO, which is veryeffective, compared with other algorithms.