In this paper, a mixed flow chemical production scheduling problem with material distribution (MFCPSP_MD) is considered. The MFCPSP_MD consists of two coupled sub-problems. One subproblem is the material distribution scheduling problem and the other is the mixed-flow chemical production scheduling problem. The optimization objective is to minimize the processing and distribution costs. This problem widely exists in the intermittent chemical industry. A hybrid reinforcement learning (HRL) method is proposed for the considered MFCPSP_MD. HRL first uses genetic algorithms to design corresponding crossover and mutation operators for the encoding scheme, obtaining an optimal elite population. Then, combined with Q-learning algorithm, it selects heuristics from a pre-designed low-level heuristic set (LLHs) to optimize the solution space and obtain better results. Considering the different types of materials required for product processes in the production process and the existence of priority relationships, a three-stage coding strategy is used and a corresponding decoding mechanism is designed to ensure the legitimacy and integrity of the solution set. Finally, the superiority of the algorithm and the validity of the model are demonstrated by solving some simulation cases.