Flow shop scheduling optimization and simulation analysis based on Q-learn algorithm and process value evaluation algorithm

Efficient and rational production flow shop scheduling (FSS) is the key to improving production efficiency, coping with market changes and promoting the core competitiveness of enterprises. In dealing with the flexible FSS problem, it has been more difficult for the traditional intelligent algorithm to find the optimal solution from the global perspective with the increase of the complexity of the system scale. Therefore, in this paper, by taking the flexible production FSS problem as the research object, the Agent-based decentralized production scheduling model guided by complex adaptive systems (CAS) was established. Besides, the model was locally optimized on the basis of Q-Learn algorithm and process value evaluation algorithm. The simulation experiment proves the feasibility of the scheduling optimization model based on Q-learn algorithm and process value evaluation algorithm in solving the flexible production FSS problem. This provides a practical solution for the actual production FSS problem. © 2018 Editura Politechnica. All Rights Reserved.