基于神经网络的电站锅炉故障诊断研究

The Diagnosis Research of Boiler Fault Based on Network

应用模糊模块化神经网络和递推合成BP网络,并结合混合型知识表示和知识获取方法、基于知识的专家系统等技术对锅炉故障诊断与预测问题进行了研究。文章所建的用于锅炉故障诊断的模糊模块化神经网络模型因进行了样本聚类,实验结果表明:其网络训练的速度和精度明显提高,同时有效地解决了BP网络应用于复杂的锅炉系统故障诊断时容易陷入局部最小点的问题。且该网络采用多输出的结构,不仅能对故障是否发生进行诊断,而且还能判断故障发生的严重程度。

The software of diagnosis and prediction of boiler fault is developed by using fuzzy modular network, recurrent composed network and the method of mixed knowledge representation and expert system technology in this paper. The simulation test result indicates that the speed and precision of sample training are increased because of sample clustering for fuzzy modular networks,the problem of slow training speed and local minimum point are avoided when BP networks are applied in the fault diagnosis of complex boiler.Multi output of the fuzzy modular networks can decide whether the fault of boiler happens,and the severity degree of boiler fault.