Parameterization Method and Constraints Transformation Method for RLV Reentry Trajectory Optimization Using Genetic Algorithm

CHEN Gang,WAN Zi-ming,XU Min,CHEN Si-lu
DOI: https://doi.org/10.3969/j.issn.1004-731x.2005.11.044
2005-01-01
Abstract:The craft reentry trajectory optimization is a kind of optimal control problem. The conventional direct and indirect methods are all very sensitive to the initial values that need much experience. Genetic algorithm is robust to initial guess, but it’s a static optimization algorithm and couldn’t be applied to optimal control problem directly. In order to design the chromosome, three parameterization methods were investigated. The constraints transformation method was also a key problem in the trajectory optimization for genetic algorithm to design fitness function. For the minimum control energy reentry trajectory with fixed terminal time, three kinds of parameterized methods and constraints transformation methods were argued in detail. The simulation results indicate the direct discrete method and dynamic penalty function method have good performance.
What problem does this paper attempt to address?