Multi-target data association algorithm based on tracking-differentiator

From the fact that position and velocity of a moving object can not vary discontinuously, a data association algorithm based on tracking-differentiator was presented for multi-target tracking in unknown clutters. The tracking-differentiator was applied to obtain positions and velocities corresponding to the measured data in target gates and multi-target data association in unknown clutter environment was realized by comparing these positions and velocities with target position and velocity at the previous moment. Since the algorithm utilizes measured data directly and does not depend on the prior statistics knowledge about target movements, sensor noise and clutter properties, it has good capability of data association in the environment with lower dense clutter and known number of targets. The algorithm possesses low computing burden and simple structure, and separates with target state filtering algorithms completely, which facilities modular design, combination with other filtering algorithms and implementation in practical engineering. Simulation results illustrate the effectiveness of the proposed algorithm.