Comparing GPU-parallelized metaheuristics to branch-and-bound for batch plants optimization

摘要

We systematically compare two approaches with the optimal design of multiproduct batch plants that are widely used, e.g., in the chemical industry. Deterministic algorithms like branch-and-bound achieve global optimality, but often require a prohibitively high computation effort. We propose an alternative, hybrid algorithm by combining two metaheuristics: ant colony optimization (ACO) and simulated annealing (SA), in order to find near-optimal solutions in a reasonable time. We develop a parallel implementation of our hybrid approach on graphics processing units using CUDA. We experimentally compare both approaches, and we show that our hybrid metaheuristic approach (ACO+SA) yields almost-optimal solutions, while computing them significantly faster than when using branch-and-bound.