Design and Trajectory Estimation of Dynamic Motion Model

Abstract

This paper illustrates linear and maneuvering motion model of target trajectories. The trajectory of the target is segmented into varied profiles considering constraints of velocity, acceleration and jerk. The influence of angular jerk, centripetal acceleration and turn radius of dynamic motion model are integrated feature of the proposed model. The mathematical motion models have been designed considering physical laws of motion and vehicle dynamics. The prior research has not focused intensively on the detailed designing of the dynamic motion model. The designed motion models are described in this paper with their vivid dynamic changes under the complexity of an environment and their adaptability to such changes. The tracking of complicated trajectories of moving target using an estimation technique based on Rao-Blackwellization integrated with dynamic motion model has been proposed in this paper as a resolution to solve illustrated tracking complexities. The performance of proposed estimation technique has been illustrated and compared with conventional tracking algorithm of Particle filter and Kalman filter. Furthermore, the proposed estimation technique is validated on the real world convoluted trajectory extracted in simulated environment of Carmaker to illustrate the performance of tracking algorithm.