本文介绍了神经元网络的感知机和反向传播算法。基于这两种模型,我们提出了一种用于故障诊断的联想搜索网络,并把它应用于一个锅炉给水装置的控制系统故障诊断上,最后给出了计算机的仿真结果。 This paper introduces the perceptron neural network and back propagation algorithm. Based on these two models, we propose a predictive search network for fault diagnosis and apply it to the fault diagnosis of a control system of a boiler feedwater system. Finally, the simulation results of the computer are given.