基于改进灰色模型的模拟电路故障预测研究

Research on Fault Prediction of Analog Circuit Based on Improved Grey Model

为保证飞行器信息处理设备供电电源模块的正常运行,对典型BUCK降压电路进行Simulink仿真模拟电容、电感退化过程,结合电容、电感理论退化模型确定供电电源模块的故障特征参数;对灰色模型建模过程进行分析,从数据的预处理和训练数据维数入手改进得到了一种具有适应度的数据平滑预处理灰色模型;等时间间隔测量信息处理设备供电电源的纹波电压值,分别用GM(1,1)模型,新陈代谢模型和改进的灰色模型对比分析,用平均相对残差作为预测精度的判断标准,验证了改进灰色模型的有效性和正确性。

In order to ensure the normal operation of the power supply module of the aircraft information processing equipment,the typical BUCK step-down circuit is simulated by Simulink to simulate the degradation process of capacitance and inductance,and the fault characteristic parameters of power supply module are determined by combining the degradation model of capacitance and inductance theory;Analyze the modeling process of grey model,from the data preprocessing and the dimension of training data,an adaptive gray model of data smoothing preprocessing is obtained;Measuring the ripple voltage of power supply of information processing equipment at equal time intervals.The GM(1,1)model,metabolism model and improved grey model are used for comparative analysis.The validity and correctness of the improved grey model are verified by using the average relative residual as the criterion of prediction accuracy.