用约束满足自适应神经网络和有效的启发式算法解Job-shop调度问题

提出一种用约束满足自适应神经网络结合有效的启发式算法求解Ｊｏｂ－ｓｈｏｐ调度问题．在混合算法中，自适应神经网络具有在网络运行过程中神经元的偏置和连接权值自适应取值的特性，被用来求得调度问题的可行解，启发式算法分别被用来增强神经网络的性能、获得确定排序下最优解和提高可行解的质量．仿真表明了本文提出的混合算法的快速有效性．

USING CONSTRAINT SATISFACTION ADAPTIVE NEURAL NETWORK AND EFFICENT HEURISITICS FOR JOB-SHOP SCHEDULING

Based on constraint satisfaction this paper proposes a new adaptive neural network, and an efficient heuristics hybrid algorithm for Job shop scheduling. The neural network has the property of adapting its connection weights and biases of neural units while solving feasible solution. Heuristics are used to improve he property of neural network and to obtain local optimal solution from solved feasible solution by neural network with orders of operations determined and unchanged. Computer simulations have shown that the proposed hybrid algorithm is of high speed and excellent efficiency.