自适应RBF神经网络在模拟电路故障诊断中的应用

Application of Adaptive RBF Neural Network in Analog Circuit Fault Diagnosis

针对模拟电路故障诊断问题,提出一种自适应RBF神经网络学习算法。该算法根据网络每次训练后得到的总体误差与上一次训练后得到的总体误差的大小关系,对传统RBF神经网络学习算法中固定的学习率和动量因子进行自适应调整,使网络学习的方向性得到增强,加快网络的学习速度。将自适应RBF神经网络应用于模拟电路故障诊断,以电压单端差分转换器电路为例进行故障诊断实验。仿真结果表明,该方法能够实现模拟电路故障的检测及定位,具有学习速度快、诊断准确率高的特点,对模拟电路故障诊断具有重要的参考价值。

For the purpose of analog circuit fault diagnosis, proposes an adaptive RBF (Radial Basis Function) neural network algorithm. According to the relationship between the total error of this training and last training, the learning rate and momentum factor in the algorithm are adjust. ed dynamically and adaptively. So, the learning direction of RBF neural network was enhanced and the learning speed is improved. This al. gorithm is used in analog circuit fault diagnosis. The single-terminal differential voltage converter circuit is taken as an example to carry out fault diagnosis experiments. The simulation results show that this method can realize fault detection and location of analog circuit. It has the characteristics of fast learning speed and high diagnostic accuracy, and has important reference value for fault diagnosis of analog circuit.