A Genetic Based Hyper-Heuristic Algorithm for the Job Shop Scheduling Problem

Job shop scheduling problem (JSP) is a typically NP-hard problem. In this paper, we proposed a genetic based hyper-heuristic algorithm for solving JSP. The hyper-heuristic algorithm employs two level frameworks. In the high-level, genetic algorithm is employed to explore the whole solution space, while in the low level, there are 16 combinations coming from 4 frequently used crossover operators and 4 mutation operators. Time window is introduced to learn from the recent searching process. Finally, the result on the benchmark instance FT10 shows that the time window helps the proposed algorithm to search effectively. Furthermore, to verify the validity of the proposed algorithm, some traditional genetic algorithms with fixed genetic operators are also employed to test FT10, supporting the statement that it is much better using adaptive operator selection in GA than using a fixed operator.