Dynamic Compensation Of Sensors Based On Improved Recursive Least Squares Algorithm

In order to eliminate the sensor's dynamic error, a method that improved recursive least squares method is proposed. Gradient descent method is used to generate the initial value of the filter parameters, then the recursive least squares method (RLS) algorithm is used to optimize the parameters. The algorithm is tested and verified on the matlab platform. The time-domain response and frequency domain response of the sensor are analyzed before and after compensation. The piezoelectric sensor CY_YD-205 is compensated. Engineering experiments show that the compensation filter designed by the improved recursive least squares algorithm can improve the dynamic characteristics of the sensor system.