摘要：AbstractControl applications in the domain of autonomous vehicles need robust, accurate and fast motion estimators. Estimation techniques generally fuse absolute and relative positions. This paper presents another approach which computes the system relative motion thanks to linear accelerometers and without rate gyroscopes. This method, known as Gyroscope-Free (GF), allows relative motion to be computed using only data from multiple accelerometers spread all around the object. A major aspect of this contribution is to get rid of any dynamic system model in the estimator such that the method proposed can be applied to any real application. A GF-Inertial Measurement Unit (IMU) relative motion estimator using an Extended Kalman filter (EKF) without system model is introduced and implemented on a real world test bench. Results show the suitability of the proposed estimator.
关键词：KeywordsEKFGyro-free IMUsSensor modelingMulti-sensor data fusionIntelligent vehicle
