电子电路模糊神经网络故障诊断研究及仿真

针对模拟电路的故障特点,在确定智能诊断算法的基础上,论文尝试用高级语言编程来仿真实现.论文主要讨论了故障诊断的神经网络方法和将输入模糊化后的模糊神经网络方法,该方法结合了模糊逻辑和神经网络的优缺点,对于电阻元件的软故障,用模糊神经网络得到了比较理想的结果,对于所选的电路,用模糊神经网络可以对电阻增大和减小进行识别,实现了模拟电路软故障的辨识.对于模糊规则分不开的故障,可以运用神经网络进一步细分.论文着重以具体的模拟电路为研究对象编程模拟了诊断方法,结果证明这样的诊断方法是有效的.

Fault Diagnosis Study and Simulation of Fuzzy Neural Network of Electronic Circuit

According to the fault characteristics of analog circuits and based on intelligent arithmetic ascertained,the paper attempts to use high-level language programming for realizing emulation.The neural network and fuzzy neural network used to diagnose analog circuits are presented in this paper.The method has combined the merits of fuzzy logic with that of neural network,and the fuzzy neural network has received perfect result for diagnosing the soft failure of resistor unit.It could recognize increased and decreased resistance for identifying the soft failure of analog circuit.If fuzzy rule couldn't distinguish failures,the neural network could do the job.The result shows that this kind of methods is effective.