Application of receding horizon control strategy to pursuit-evasion problems

Time-optimal interception of an arbitrarily manoeuvering evader is investigated using a simplified kinematic model of planar constant speed motion. Against such an uncertain evader behaviour a new and robust pursuer strategy, inspired by the 'receding horizon' control concept, is developed. Numerical results of an extensive simulation study show that for all the cases that have been tested the optimal horizon length is approximately 60% of the minimum time required to capture a straight-flying evader. Moreover, the performance loss compared with the optimal control and game solutions is less than 1%. The proposed guidance algorithm, based on the 'receding horizon' control concept and using a singular perturbation approximation, is very simple to implement and leads almost always to a near-optimal outcome.