Tracking algorithms commonly use some practical models of target motion to estimate the present and the future target kinematics quantities such as the position, the velocity and in certain cases, the acceleration. When there is a maneuver, the tracking algorithm will detect the error created by the change in target motion and correct the situation to adapt itself to this new change or new tracking model. There are different approaches in the literature for handling maneuver detection such as parallel filtering techniques. These techniques are used mainly because of quick response in maneuver detection of moving objects and to enhance the position and the speed estimations with filtering stability. The purpose of the study is to show the benefit of using a parallel filtering technique over a single filter. The paper presents a parallel filter design using three extended Kalman filters (EKF) with a simple switching algorithm for maneuver detection. The state vector of the EKF is in cartesian coordinates and the measurement vector is in polar coordinates. This design is relatively simple compared to other parallel Kalman filter techniques and requires modest computer resources. The simulation results have shown improvement using parallel filtering, particularly in smoothing speed estimations and the rapidity of convergence to track the target after it has abruptly maneuvered.
展开▼
