Study On Low-thrust Trajectory Design And Optimization Method For Deep Space Explorer And Application

In the new wave of deep space exploration, low-thrust propulsion technology attracts widespread attention and develops rapidly. It is represented typically by the electric propulsion system, which has a high specific impulse and therefore low fuel consumption. Under this background, the low-thrust trajectory design and optimization method and its application to deep space exploration is studied in this thesis. The main contents of this thesis are as follows.The Gauss pseudo-spectral method is investigated. And it is improved in two aspects, i.e., a quick search of the mission window and initial guess generating. Gauss pseudo-spectral method is a common method for low-thrust trajectory design. It uses global Lagrange interpolation polynomials to fit state and control variables. Using Gauss pseudo-spectral method, the low-thrust optimal control problem is converted into a nonlinear programming problem. In low-thrust trajectory design, departure time and flight time has large range of span and strong discontinuity. They have significant impacts on design results. To give a prior search of these mission window parameters, a method based on multi-impulse energy estimation is proposed. Meanwhile, a new way to quickly generate an initial guess for Gauss pseudo-spectral method is represented, which converts a low-thrust trajectory directly from a multi-impulse one. The generated initial guess plays a catalytic role on improving the robustness and computational efficiency of Gauss pseudo-spectral method.To overcome the technique bottleneck that the numerical integration of low-thrust trajectory is inefficient in direct-shooting method, a dynamics decomposition method is proposed. Referring to the Multi-Conic technique for computing Earth-Moon transfer, the dynamic decomposition method decomposes the trajectory into three relatively independent motions during a time interval. The terminal state of the interval is obtained analytically by the synthesis of these three motions. Finally, the entire trajectory is calculated rapidly by a recursive procedure. The trajectory accuracy and computational efficiency of dynamic decomposition method is analyzed and validated through extensive simulations. The dynamic decomposition method is also compared with the improved Gauss pseudo-spectral method.In order to simplify the process of low-thrust trajectory design, a novel method based on virtual gravitational center(VGC) is proposed. The VGC is assumed to have exact the same gravitational constant with the Sun. The spacecraft’s motion, subject to the Sun’s gravity and low-thrust, is converted into the VGC’s motion. It can be shown that the low-thrust trajectory is still a conic arc when expressed in a translational reference frame centered at the VGC. Thus, the low-thrust trajectory design is transformed into the design of VGC’s motion path. Numerical results show the equivalent between the low-thrust trajectory and the VGC’s motion path. A comparative analysis on three different implementations of the VGC based method and the dynamics decomposition method is given.The global optimization scheme of multi-target exploration and its sequence selection method is studied. It is tended to visit several targets in a deep space exploration project to increase the scientific return. This turns out to be a global optimization problem with multiple objectives. To solve this problem, the global optimization scheme of multi-target exploration, which consists of four related parts, is given. And then, two multi-target exploration sequence selection methods, i.e., flow tube selection method and backtracking best-first search method, are proposed. Finally, the global optimization scheme and two multi-target exploration sequence selection methods are validated and analyzed through a typical low-thrust multiple asteroids exploration problem.Low-thrust trajectory design methods under different ideas and principles are studied systematically in this thesis. The global optimization scheme of multi-target exploration and its available sequence selection method is also proposed. The methods and results presented in this thesis have evident reference meanings for future deep space exploration of our country.