A hybrid optimization algorithm for the job-shop scheduling problem

Qiang Zhou, Xunxue Cui, Zhengshan Wang, Bin Yang

Abstract: The job-shop scheduling problem is a NP-hard combinational optimization and one of the best-known machine scheduling problems. Genetic algorithm is an effective search algorithm to solve this problem; however the quality of the best solution obtained by the algorithm has to improve due to its limitation. The paper proposes a novel hybrid optimization algorithm for the job-shop scheduling problem, which applies chaos theory on the basis of combining genetic programming and genetic algorithm. It improves the quality of the initial population by using chaos optimization method; it maintains the population diversity by chaotic disturbance and anti-equilibration in crossover of genetic programming. Three traversals are adopted to reduce the chance of reaching local optimal solution. Moreover, a scheme of changing weight is proposed during the process of evolution to increase the global exploration capability. The experimental results show that the effectiveness and good quality of the hybrid algorithm is obvious from some benchmarks.