Chaotic multi-population ABC algorithm based on memory and levy flight for solving dynamic job shop scheduling problems

In the real world, most of the problems are dynamic optimization ones. In other words, optima may change over time. Algorithms that can solve these kinds of problems can adapt well, by using the ability to track the optima in case of evolution. In this paper, a novel chaotic multi-population artificial bee colony optimization with levy flight algorithm (CMABCLA) is proposed to minimize makespan in dynamic job shop scheduling problem. The dynamic events considered in this paper are machine breakdowns, random job arrivals and processing time changes. Dynamic job shop scheduling is a known NP-hard combinatorial optimization problem. The chaotic system used in this algorithm has more precise prediction of the future than the random system, and it increases the convergence rate of the algorithm. Also, after a change, the information obtained from the previous state makes quick adaptation possible. Moreover, the utilization of levy flight in the scout bees phase has led to the improvement of exploration. Moving Peaks Benchmark has been chosen to examine the effectiveness of the proposed method. The results of conducted experiments show the superiority of the proposed method to state-of-the-art algorithms in terms of offline error and CPU time.