An algorithm of real-time solving deadlock for job-shop schedule

Genetic algorithm is widely applied for the Job Shop scheduling Problem (JSP) and is proved to be a better solution compared with most conventional solutions, however, the general methods to finding optimal solution always abandon the deadlock chromosomes. Two different methods for coding are compared in this paper. On this basis, a novel algorithm with real-time discovery and solving the deadlock is presented, which does not abandon any chromosomes and just adjusts the genes' dispatching orders in deadlock chromosomes. It schedules all chromosomes, and finds out the optimal solution quickly. Simulation experimental results show this algorithm is effective.