Heuristic Particle Swarm Optimization for resource-constrained project scheduling problem in chemical industries

This paper considers the resource-constrained project scheduling problem (RCPSP) with features on batch scheduling in the chemical industries such as multi-product facilities, set-up time depending on sequence, combination of divergent and convergent production flow, recycle of material. For this complicated problem, we present a Heuristic Particle Swarm Optimization (HPSO) where some strategies including Batch Splitting Mechanism (BSM), Stochastic Precedence Search Scheme (SPSS), Improved Schedule Generation Scheme (ISGS) and Recycle Material Scheme (RMS) are proposed to improve the HPSO. Computations show that 68% solutions from HPSO are equal to or better than the best solutions known so far. And compared with the best algorithms on this problem in the literature, Decomposition Approach (B+BS) and Time Grid Heuristic (TGH), HPSO improves 6.23% over B+BS and 37.64% over TGH as far as average deviation is concerned.