电站锅炉神经网络燃烧诊断系统应用研究

先进的燃烧诊断技术可以有效地提高电站锅炉运行的经济性和安全性，本文通过对炉膛火焰的图像采集，利用计算机数字图像处理技术及人工神经网络模型分析方法，开发了永安电厂5号炉的火焰图像燃烧诊断系统。该系统为运行人员提供了有意义的定量化特征参数，并对燃烧状况辩识的机理进行了富有成效的探索，为电站锅炉的燃烧诊断和优化控制提供新方法和新途径。

Applied Research of a Neural Network-based Combustion Diagnostic System for a Utility Boiler

The use of an advanced combustion diagnostic system can be conducive to an effective enhancement of utility boiler operation economy and safety. Through the collection of furnace flame images and by utilizing computer based digital image processing techniques as well as the analysis method of an artificial neural network model a combustion diagnostic system of flame images has been developed for boiler No. 5 of Yongan Thermal Power Plant in Fujian Province. This system has provided meaningful quantified characteristics parameters, performing a highly effective probing of the combustion condition identification mechanism. As a result, a new method and approach for combustion diagnosis and optimized control is provided for utility boilers.