神经网络在城市综合业务故障诊断中的应用

Application of Neural Network to City Integrated Service Fault Diagnosis

提出计算机城市综合业务实时处理网络系统中的在线故障诊断问题,分析了计算机城市综合业务中应用神经网络技术进行故障诊断的可行性.重点针对应用软件故障,给出了后台服务中常见的故障征兆原因关系例图、故障诊断神经网络模型、神经网络参数和学习样本.通过仿真实验,验证了BP神经元网络能够有效实现后台管理中的故障诊断.

This paper proposes on-line fault diagnosis issues in the computer city intergrated service real-ime processing network system, analyzes the feasibility of applying the BP nerve cell technique to computer city integrated service transaction for fault diagnose. It mainly aims at the malfunctions of application software, presents some relationship diagrams to illustrate reasons of common background fault symptoms, the model of fault diagnosis nerve network , the nerve network parameters and study samples. Finally, proves the validity of BP nerve cell network technique in background management is proved by simulation experiment.