关键词: 差分进化算法, 车间调度, 调度算法, 不确定性, 流程工业

Abstract: Considering uncertain factors in actual production scheduling of process industry, such as Input-output ratio, equipment conversion, and so on, the uncertain process shop scheduling model was established to maximize area of satisfaction based on fuzzy theory. In order to improve optimization performance of algorithm, a multi-variant and two-way cross differential evolution algorithm was presented to increase population of individuals in search space traversal and merit-based selection. Finally, the improved algorithm was applied to solve polyvinyl chloride resin shop scheduling of an electro-chemical plant, and the influence of time length, different mutation strategies and initial capacity of intermediate storage on scheduling results were analyzed. Simulation results revealed effectiveness and stability of the improved algorithm.

Key words: differential evolution algorithm, shop scheduling, scheduling algorithms, uncertainty, process industry