一种基于神经网络的蓝牙汽轮机故障诊断装置

A bluetooth turbine fault diagnosis device based on neural network

在发电厂中,伺服模件控制着汽轮机蒸气阀门的进气量,驱动转子为发电机提供动力,是汽轮机系统的核心部件。在传统汽轮机故障诊断中,往往需要依赖工程师的经验来定位故障,但由于部件的老化故障早已出现,人工分析往往匮乏充分数据无法判断。通过在伺服模件上加装蓝牙模块,利用蓝牙5.0 Mesh技术采集多个伺服模件的关键数据,并通过神经网络技术对数据和诊断结果进行训练,构建出一个可用于电厂汽轮机关键设备故障诊断的智能诊断装置。

In plant as core parts servo modules control intake volume of steam valve of turbine to drive rotor and power motor.In traditional turbine fault diagnosis locating fault relies on engineers’experience.However,because aging failure of parts has already occurred long ago,it is impossible to be judged by manual analysis of lacking data.In this paper through the method of adding BLE module on servo modules and sampling key data of servo modules by BLE 5.0 Mesh technology,I use ANN(Artificial Neural Network)technology and create a smart fault diagnosis device of analyzing key equipment failure of turbine by training input data and diagnosis result in ANN.