Novel Strategy For Spacecraft Close Range Proximity With Obstacle Avoidance Considering Uncertainties And Arbitrary Shape

With the development of aerospace technology,the spacecraft close-range proximity with obstacle avoidance has attracted much attentions in recent years.In this thesis,the spacecraft close-range proximity with obstacle avoidance has been studied for the application scenarios.Subsequently,the influences of the uncertainties and the arbitrary shape are analyzed and the corresponding avoidance strategies are developed.Furthermore,the improved LQR controllers with respect to the novel avoidance strategies are designed and is applied to track the reference trajectory.The main work of this thesis includesFirstly,in terms of the spacecraft close-range proximity with obstacle avoidance and uncertainties,the equal-collision-probability-curve method(ECPC)is proposed.Compared to traditional collision probability,the ECPC method does not contain transcendental elements and hence the computational burden can be greatly decreased while maintaining the necessary accuracy.Moreover,analytical validation is performed to assess the use of such collision avoidance scheme for safety critical operations.However,the ECPC method models the target spacecraft as spheres and the complex shape of the target spacecraft deteriorates the safety performance.Thus,the multi-equal-collisionprobability-curve method(MECPC)method and the dual-equal-collision-probability-curve method(DECPC)method are proposed to solve the obstacle avoidance problem in the presence of complex shape.Subsequently,based on the novel avoidance strategies,the corresponding improved Linear Quadratic Regulator(LQR)controllers are designed to track the reference trajectory and a Lyapunov-based analysis verifies the stability of the overall closed-loop system.Then,to solve the fast motion planning problem with collision avoidance of arbitrary shape obstacles,a novel Gaussian Mixture Model(GMM)strategy is proposed.Firstly,using the K-means and Expectation Maximization(EM)algorithm,the statistical representation of the model's geometrical shape is obtained.Next,similar to the model construction of the Gaussian Mixture Model,the novel GMM-based potential function is established.Compared with the traditional potential function,the available parameters of the complex shape are obtained and included in the proposed function.Thus,the influence of the geometrical shape is considered.Moreover,the corresponding improved LQR controller with respect to the GMM method is designed to track the reference trajectory and a Lyapunov-based analysis verifies the stability of the overall closed-loop system.Furthermore,based on GMM method,the adaptive Gaussian Mixture Model(AGMM)is proposed to consider the joint influence of the uncertainties and the arbitrary shape.Finally,in terms of the motion planning problem of spacecraft proximity operations with obstacle avoidance under limited uncertainty,the improved equal-collisionprobability-curve(IECPC)strategy is proposed.Combining the collision probability function method,AGMM method and IECPC strategy,the safety zone(SZ)method is proposed to solve the safe close-range proximity operations in the presence of the uncertainties and the arbitrary shape.Compared with the ECPC method and the IECPC strategy,the SZ method can considering the influence of the arbitrary shape.Furthermore,compared with the GMM method and the AGMM method,the SZ method can reduce the collision probability.Moreover,the corresponding improved LQR controller with respect to the SZ method is designed to track the reference trajectory and a Lyapunov-based analysis verifies the stability of the overall closed-loop system.In conclusions,the numerical simulations verify the effectiveness of the proposed avoidance scheme.Compared with the traditional methods,the proposed methods can solve the obstacle avoidance problem in the presence of the uncertainties and the arbitrary shape.Thus,the proposed methods have broad prospects for engineering applications.