A robust method to determine relevant target vehicle using vehicular radar

In order to solve the problem of the instability in the selection of relevant target, caused by the complexity of the measurement environment of vehicular radar, a robust method to determine the relevant target using vehicular radar is proposed. Based on analyzing the measurement environment, the method uses the principle of the nearest object in the same lane for target pre-selection. The Kalman filter is applied to predict the target information and the relevance verification of the pre-selected target is done by the consistence checking. The target decisions are made through a "relevant target life cycle" method. The verification tests show that by efficiently eliminating the effects of ghost objects, other disturbances and the bumping and swinging of vehicle, the proposed method can accomplish the determination of relevant target under different conditions.