The hot-rolling batch scheduling method based on the prize collecting vehicle routing problem
Tao Zhang, Wanpracha Art Chaovalitwongse, Yuejie Zhang, Pãnos M. Pardalos · Journal of Industrial and Management Optimization · 2009
This paper studies a hot-rolling batch scheduling problem, which isa challenging problem commonly arising in the iron-steel industry.This problem deals with forming steel strips into rolling units anddetermining the rolling sequences while minimizing changes incharacteristics (e.g., width, thickness and rigidity) of allneighbor steel strips and maximizing the machine utilization. Basedon technical rules used in iron-steel production practice, weformulate this problem as a prize collecting vehicle routingproblem, which is a hard combinatorial multi-objective optimizationproblem. We develop a new heuristic approach to solve this problemby enhancing the framework of particle swarm optimization (PSO). Thekey features of our approach are the utilization of number mappingfunction, which allows the PSO algorithm to deal with discreteproblems, and the employment of tabu search at the end of every PSOiteration. We investigate and measure the performance of theproposed approach using three real life datasets obtained from awell-known iron-steel production company in China. The resultssuggest that our approach is very efficient and effective inproviding high-quality and practical solutions, and our approachoutperforms traditional PSO and tabu search algorithms based onthese datasets.