Hybrid optimal trajectory generation using genetic algorithm and sequential quadratic programming

The maximum performance is required for the space plane which will be realized in future. Thus, the efficient real-time calculation method for the trajectory optimization is required. The Space Systems Laboratory of Kyushu Institute of Technology has been studying the winged rocket development program as a research subject of fully reusable space transportation systems. This paper proposes an optimization trajectory method for winged rocket aiming at onboard real-time calculation. The present optimization employs a technique to solve the equations of motion directly with the control input function, which has variable coefficients for optimization. As the results, the optimization variables are reduced and the initial conditions are satisfied. GA (Genetic Algorithm) provides an initial solution of variable coefficients of the control input function for SQP (Sequential Quadratic Programming) method. This optimization algorithm using GA for initial solution and SQP for fine solution reduces the total calculation time and improves the accuracy.