Genetic Programming-based Heuristic Generation Algorithm for Steelmaking - Continuous Casting Scheduling

This paper deals with the special mixed flow shop scheduling problem of steelmaking and continuous casting, establishing the SCC production scheduling solution framework based on charge sequencing rules and device assignment rules, and proposing a heuristic generation algorithm of SCC based on GP. The algorithm can automatically generate scheduling heuristic rules suitable for production scenarios. The simulation experiment is carried out with the actual production data of one refining steel mill, The experimental results show that the scheduling rules generated by the heuristic generation algorithm based on GP have better performance than the benchmarking rules in actual production environment, and with the increase of examples, the performance advantage of generated rules are more obvious and stable.