A deep reinforcement learning algorithm for order acceptance decision of individualized product assembling

Aiming at the order acceptance decision problem of modular production enterprises, a semi Markov decision process-based order acceptance decision model is constructed. The enterprise's production capacity is divided into four independent production units, and each production unit has its daily capacity and total capacity limit. The deep reinforcement learning algorithm is used to explore and learn in the model to maximize the overall revenue when the production capacity and order delivery time are limited. At the same time, it also solves the problem that the state space is complex that the traditional Q-learning algorithm is not competent. The strategy's effectiveness is verified by comparing the trained strategy with First-Come-First-Serve(FCFS) strategy in the same environment.