基于人工神经网络的锅炉火焰温度场测量方法

讨论了用辐射测温方法测量高温煤粉炉温度场时存在的问题,提出了一种基于BP网络模型的温度场测量方法.利用设计的一套测量系统对某200 MW的锅炉的燃烧过程进行了试验,结果表明,与传统的比色法测量相比,该方法具有误差小、测量精度高、适应性强等优点,可满足系统燃烧诊断和实时测控的要求,工程上有一定的应用前景.

Measuring method of the high temperature field of boiler flame based on neural network

In accordance with the problem existing in the temperature field measurement of high temperature powdered coal boiler based on radiation method,a boiler temperature field measuring method based on B-P neural network is put forward.A test on a 200 MW boiler is made using the designed measurement system.The result of the test shows that,the measurement method has higher measuring accuracy and better suitability than traditional colorimetric measuring method.Its relative error is less than 2.3%,which is far less than that of colorimetric measuring method.The technology may be used for the diagnosis and real-time monitoring of boiler burning,and it has good application prospects.