Hybrid Evolutionary Algorithm for the Optimization of Interplanetary Trajectories

摘要

A hybrid evolutionary algorithm is applied to the design of interplanetary trajectories with multiple impulses andngravity assists. The optimization procedure runs three different optimizers based on genetic algorithms, differentialnevolution, and particle swarm optimization “in parallel”; the algorithms, which can also be employed separately, arenused synergistically here by letting the best individuals, found by each algorithm, migrate to the others at prescribednintervals. A comparison with the results presented in recent literature, which state that differential evolution is wellnsuited to deal with this kind of problem, is carried out. The performance of the hybrid optimizer is comparable to thatnof differential evolution in terms of computational time and function evaluations when problems with a reducednnumber of variables are considered. The hybrid optimizer may instead exhibit better performance when morencomplex problems are dealt with. The results also show that the algorithm performance is remarkably improved bynintroducing a “mass mutation” operator to avoid premature convergence to suboptimal solutions and by means of anparticular choice of the variables to describe the trajectory.