Optimal Tracking Gate Based on Hybrid Encoding Genetic Algorithm

The track qualities can be depended on tracking algorithm optimization.One of the essential optimization problems in tracking is gate selection.In this paper,a hybrid encoding genetic algorithm is proposed to off-line optimization of tracking gate parameters for maneuvering target tracking in clutter.Binary string and floating-point string represent shapes and size of tracking gate respectively.Hellinger distance is chosen for performance evaluation of tracking and can be core part of the fitness function of genetic algorithm.Generally speaking,the tracking system optimization can be converted into genetic algorithm optimization,then the tracking gate parameters can be efficiently tuned in different scenarios.