The concept of relative priority is introduced in this paper to attempt to minimize job lateness for a class of job shop scheduling, where the throwing time and delivery date are considered. A modelling methodology is proposed to illustrate the class of job shop, including its static and dynamic characteristics. The possible position of a job in a schedule is restrained by its effect on objective function value. Consequently the state space to be searched can be greatly reduced and its quality can also be improved so that it is easier and quicker to step to the minimization point. Detailed definition of relative priority is given in this paper and algorithms are discussed how to get a stable relative priority set. For both static case and dynamic case, it is possible to obtain a solution closer to optimum in the similar way. Finally data experiments show the feasibility and efficiency for proposed relative priority approach.
展开▼
