Predictive pursuit-evasion game control method for approaching space non-cooperative target

This paper designs the predictive pursuit-evasion game-based orbit control method for the chasing spacecraft to approach a space uncooperative maneuvering target. Firstly, the trajectory planning algorithm combines RRT∗ and cubic splines to generate a feasible reference trajectory in the relative motion space for the chaser, which enables to deal with the boundary constraints and avoid collisions with the attachments of target. Then, based on the dynamics model, input constrains and objective functions, the pursuit-evasion game model between the target and the chaser is formulated, in which each one acts independently to satisfy its own objective function. With the game formulation, a predictive pursuit-evasion game controller for the chaser is designed to track reference trajectory. Under the frame of model predictive control, the multiple objective constraint optimization can be transferred into to quadratic programming problem to handle input constraints. By predicting opponent's best move and changing its optimal strategy for its benefit iteratively, the saddle points of the game can be obtained without opponent's maneuvering. Numerical simulations verify the effectiveness of the predictive pursuit-evasion game control method for approaching an uncooperative target.