Design and implementation hardware in the loop for soft landing on Mars

Abstract

In this thesis an optimal control problem dealing with soft landing on Mars is studied. One of the important issues in Mars or moon landing is to guide the surface lander in such a way that it reaches to a specified position with zero contact velocity. Referred as soft landing, this problem has obviously different solutions without imposing a performance index. One of the most useful performance indices one may apply here is fuel consumption. Due to requiring fuel for coming back to the parking orbit after mission, it will be useful if we could control the lander to the Mars surface with minimum fuel consumption. Therefore, the problem is modeled as to control a time-varying dynamical system from a given position and velocity to a desired state with zero contact velocity. Because of the time varying coefficients in differential equations, the analytical methods fail to find a solution and there are interests in numerical methods. For such an applicable problem that the initial conditions are unknown before launch and they are determined in descent stage, the required time for solving the problem is very critical. In this thesis, the method of iterative dynamical programming (IDP) is used to solve the problem. Once the suitable initial conditions for landing are determined, the flight computer will find the control commands for thrusters in a reasonable time. To check the efficiency of the method we use hardware in the loop simulation (HITLS). A graphical user interface is developed to set the value of initial conditions, to see the resulting effect on the Mars surface lander in a visual environment and to compare the fuel consumption in two cases where optimal control is used and is not used. Case studies and HITLS show the efficiency of the proposed method for driving control commands. The method is also implemented parallel on flight computers with multiple microcontrollers to reduce the time of running IDP subroutine as well.