基于神经网络的实验台温度控制系统故障诊断研究

Research on Fault Diagnosis about experiment platform temperature control system based on Neural Network Theory

实验台温度控制系统工作的正常与否是决定煤自然发火实验成功与否的关键因素,故进行实验台温度控制系统的故障诊断至关重要。本文在构建温度数据采集处理系统的基础上,利用神经网络理论中的BP网络对温度控制系统进行故障诊断,构建三层BP网络结构,设定阈值,建立故障诊断软件,并采用MATLAB6.1进行实现。以实例说明BP网络具体训练结果,故障诊断效果较好,对实验台的正常运行具有较重要的指导作用。

The working order of experiment platform temperature control system is the pivotal factor to decide the working order of coal autonomic combustion. So it is very important to carry through fault diagnosis of temperature control system. In this article , based on building data collection system, we process fault diagnosis about temperature control system using BP network of neural network system, form three-layered BP network structure, enactment threshold value, design software on fault diagnosis and carry it out with MATLAB 6.1. With instance the result of BP network studying is given, and the purpose of fault diagnosis is good, so it is important to direct the natural function of the experiment platform.