《An improved constraint satisfaction adaptive neural network for job-shop scheduling》.pdf

J Sched (2010) 13: 17–38 DOI 10.1007/s10951-009-0106-z An improved constraint satisfaction adaptive neural network for job-shop scheduling Shengxiang Yang · Dingwei Wang · Tianyou Chai · Graham Kendall Received: 30 April 2007 / Accepted: 6 February 2009 / Published online: 21 March 2009 © Springer Science+Business Media, LLC 2009 Abstract This paper presents an improved constraint sat- sis of many state-of-the-art scheduling systems. Hence, it isfaction adaptive neural network for job-shop scheduling may also be used to construct advanced job-shop scheduling problems. The neural network is constructed based on the systems. constraint conditions of a job-shop scheduling problem. Its structure and neuron connections can change adaptively ac- Keywords Job-shop scheduling · Constraint satisfaction cording to the real-time constraint satisfaction situations that adaptive neural network · Heuristics · Active schedule · arise during the solving process. Several heuristics are also Non-delay schedule · Priority rule · Computational integrated within the neural network to enhance its conver- complexity gence, accelerate its convergence, and improve the quality of the solutions produced. An experimental study based on 1 Introduction a set of benchmark job-shop scheduling problems shows that the improved constraint satisfaction adaptive neural net- The job-shop scheduling problem (JSSP) is one of the most work outperforms the original constraint satisfaction adap- difficult problems in scheduling. It aims to allocate a num- tive neural network in terms of computational time and the