应用自适应滤波抑制动态测试随机误差

Applying Adaptive Filtering to Reduce the Random Error in Dynamic Measurement

本文提出应用递推预测误差辨识与卡尔曼滤波相结合的自适应滤波方法,来抑制动态测试的随机误差,以给出更为精确而可靠的测量结果。同时,提出对数据采取中心平滑作预分解处理,避免较大趋势项的影响,并通过大量仿真试验给出不同情况下模型阶数及各算法参数的合理取值。该方法可抑制动态测试随机误差约30%～50%,且要求先验信息较少,适用性广,计算速度快,可用于实时处理。

An adaptive filtering method is used in this paper,which combinesthe recursive prediction error identification with the Kalman filtering,to reduce the random error in dynamic measurement,so as to obtain more precise and reliable result of a measurement.In the meantime,in order to avoid the influence of the notable trend of data,the moving average method is used to predecompose the data.A series of simulating tests is also completed to get the suitable value of the model order and other parameters of algorithm in various cases.The results of many simulating tests and applied examples show that the variance of dynamic measurement random error can be cut down about 30%~50%.By this method only a little prior informations are needed and the calculation is rapid,so it can be used in real time process.