基于改进BP算法的煤灰结渣特性诊断模型

A Model for the Diagnosis of Coal Ash Slag-buildup Characteristics by Using an Improved BP Algorithm

在传统BP算法的基础上对BP算法进行了改进 ,并引入了共轭梯度和自适应学习速率 ,其收敛速度快 ,能够有效避开局部极小值 ;以实际观测数据为样本 ,应用改进的BP算法建立了一种新的煤灰结渣特性诊断模型 ,结果表明该诊断模型可靠性高 ,具有重要的实用价值。

An improved BP algorithm is proposed based on a traditional BP algorithm. A conjugate gradient and a self adaptive learning rate have been introduced into the proposed algorithm. With a high speed of convergence it can effectively evade local minimum values. By taking actually observed data as samples a new model for the diagnosis of coal ash slag buildup characteristics was set up through the use of the improved BP algorithm. It is found that the diagnostic model features high reliability and is of high practical value.