Optimizing integrated lot sizing and production scheduling in flexible flow line systems with energy scheme: A two level approach based on reinforcement learning

Abstract

Many production environments are faced with the need to simultaneously determine the planning of lot sizing and the scheduling of production sequences while ensuring cost minimization. This issue becomes even more complex when integrating multiple energy sources with the goal of a low-carbon economy. To address this challenge, this paper proposes an integrated lot sizing and flexible flow line production scheduling model under a time-of-use pricing scheme. In addition, the model takes into account conventional grid power, on-site renewable energy sources, and an energy storage system. The associated objective function is solved adopting a two-level approach to optimize energy costs while maintaining production throughput and meeting customer demand. The implementation relies on reinforcement learning capabilities to tackle complexity issues. The proposed approach is evaluated on a benchmark case and its results are compared with those obtained with First-In-First-Out heuristic, genetic algorithm and CPLEX. These results highlight the promising aspect of the proposed approach in terms of performance.