Experimental study of a novel filter structure designed for MEMS-based sensors in electric vehicles

Abstract

This work describes a comparative study between Kalman filter, a complementary filter and a combination of both, for use in electrical vehicles. Combining the benefits offered by each filter to obtain an optimised filter combination is targeted. Three different combinations: The Kalman-complementary filter (KCF), complementary-Kalman filter (CKF) and 2KCFs are examined here. The filters are used to improve signals obtained via two sensors (gyroscope and accelerometer) integrated into the sensor IMU-MPU6050, with internal DMP. The sensor data are filtered to guarantee the movement quality of electrical vehicles. The KCF combination shows higher performance than the CKF combination. Moreover, the experimental results show that the 2KCF combination yields best performance with minimal noise levels and more accurate angle measurement. The optimal combination is strongly recommended for future electrical vehicle development.