一种基于模糊示例学习的煤粉燃烧状态诊断方法的研究

Study on a method in pulverized coal combustion diagnosis based on learning from fuzzy examples

根据火焰图像提供的火焰黑度及辐射强度等信息,基于示例学习的模糊ID3算法,给出了一种煤粉燃烧状态诊断的新方法,并提出了折线形隶属函数的概念,通过修正隶属函数,可以明显提高诊断的准确性。诊断结果表明这个方法是可行的。

A new method in pulverized coal combustion diagnosis was proposed based on fuzzy ID3 algorithm in terms of the emissive and radiation intensity of flame. The idea of broken line membership function was presented; the accuracy of combustion diagnosis was improved by correcting the membership function. The diagnosing results show that the new method is feasible.