摘要
容差模拟电路故障的多样性使得神经网络训练样本数量增加,BP网络结构趋于复杂,训练速度降低。针对反向传播神经网络(BPNN)学习收敛速度慢、易陷入局部极小值等问题,提出了基于概率神经网络(PNN)的容差模拟电路故障诊断方法,与传统的BP网络模型相比,该方法具有训练时间短且不易收敛到局部最小的优点。仿真实验表明:诊断过程快速,结果准确而且对软故障也有较高的识别能力。
The variety of faults in analog circuit with tolerance makes the number of training samples of neural network greatly increase. Structure of BP network tends to be complex and training rate is greatly reduced. Against the shortcomings of Back-propagation Neural Network （BPNN）,which include slow learning speed of convergence and the nature which is easy to fall into local minimum value, PNN based diagnostic method for faults of analog circuit with tolerance is proposed. Compared to the traditional network model of BP, it has advantages of short training time and difficult converging to local minimum. Simulation results show that： the diagnostic method is rapid, accurate and also have higher recognition ability for the soft faults.
出处
《微计算机信息》
北大核心
2008年第34期145-146,152,共3页
Control & Automation
基金
基于多元信息智能融合的模拟电路在线故障诊断研究
颁发部门:国家自然科学基金委员会(60673084)
基于多层融合模式的模拟电路智能故障诊断方法研究
颁发部门:湖南省自然科学基金委员会(06JJ3075)
关键词
概率神经网络
容差
故障诊断
蒙特卡罗分析
Probabilistic Neural Networks
Tolerance
Fault Diagnosis
Monte Carlo Analysis
