关键词: 车间调度, 随机规划, 多目标优化, 不确定信息

Abstract: To make the correct executive decision under uncertain job-shop information environment, a stochastic programming model supporting multi objective and multi priority for job-shop scheduling was proposed and a hybrid intelligent algorithm consisting of three steps was designed to solve this problem. Firstly, some approximate data samples were obtained by stochastic simulation. Secondly, a neural network model was constructed to approach the uncertain function of objectives and constraints. Then the genetic algorithm was used to search for the optimal solution. Finally, a case study of a scheduling problem in a job-shop for die manufacturing in a motor company was used to illustrate the feasibility and practicability of this model and algorithm.

Key words: job-shop scheduling, stochastic programming, multi objective optimization, uncertain information