Optimization of Filter Parameters for a MEMS IMU Accelerometer Using an Artificial Immune System

A new filter parameter optimization approach is proposed for accelerometers. Filter parameters are optimized by comparing accelerometer data with experimental acceleration values using an artificial immune system (AIS)-based algorithm. The cut-off frequency and the order of the filter can be determined in different acceleration ranges depending on the dataset. A Butterworth low-pass filter is used in simulations, but the proposed approach can easily be extended to the other filter types. Experimental test results confirm that the proposed algorithm can reach the minimum absolute mean error and the global minimum after a few iterations.