In a real manufacturing environment, the set of tasks that should be scheduled is changing over the time, which means that scheduling problems are dynamic. Also, in order to adapt the manufacturing systems with fluctuations, such as machine failure and create bottleneck machines, various flexibilities are considered in this system. For the first time, in this research, we consider the operational flexibility and flexibility due to Parallel Machines (PM) with non-uniform speed in Dynamic Job Shop (DJS) and in the field of Flexible Dynamic Job-Shop with Parallel Machines (FDJSPM) model. After modeling the problem, an algorithm based on the principles of Genetic Algorithm (GA) with dynamic two-dimensional chromosomes is proposed. The results of proposed algorithm and comparison with meta-heuristic data in the literature indicate the improvement of solutions by 1.34 percent for different dimensions of the problem.
An improved deep reinforcement learning approach for the dynamic job shop scheduling problem with random job arrivals