Partical filtering for orientation determining using inertial sensors IMU

摘要: Using a low cost Inertial Measurement Sensor (IMU) to estimate an accurate orientation of human body is challenging work. Several approaches has been proposed minimize the sensor error as Particle filter (PF) which gain popularity with nonlinear and non-gaussian noise system models. In case estimation positioning PF employed in frame hybridization other measurement systems such visual systems, radio-localisation, GPS etc. this study, we propose use only data obtained from inertial sensor, constrained PF. The Iterative Mean Density Truncation algorithm (IMeDeT) quaternion based on acceleration magnetic field. An improvement term accuracy clearly detected well convergence, against Extended Kalman Filer (EKF)and optical system.