论文部分内容阅读
介绍了利用神经网络和模拟退火技术来求解有约束的ＦＭＳ资源调度问题的方法，有约束的ＦＭＳ资源调度问题首先被分解为一系列时间间隔的调度，这些时间间隔的调度由事件驱动，随着这些时间间隔的调度的完成，整个调度过程结束。仿真结果表明，由于这种方法是梯度下降法和随机搜索法的综合，因而它可以克服通常神经网络容易陷入局部极小点的缺点，从而获得问题的最优解
The method of solving constrained FMS resource scheduling problem by using neural network and simulated annealing technique is introduced. The constrained FMS resource scheduling problem is first decomposed into a series of scheduling with time interval. The scheduling of these time intervals is driven by events The completion of the scheduling of these time intervals, the entire scheduling process is over. The simulation results show that this method can overcome the shortcoming that normal neural network can easily fall into local minimum because of the combination of gradient descent method and random search method so as to obtain the optimal solution of the problem