Spatial Target Access Optimization Technology Based On Learning Evolution

In recent years,in-orbit service technology,as an emerging technology,has attracted more attention,and has been applied frequently in the maintenance of space objects in orbit(such as spacecraft and space debris).In principle,any kind of in-orbit service is a space target rendezvous process of orbital planning and optimization,that is,the service spacecraft performs orbital maneuver and planning,approaches the target spacecraft,and then completes in-orbit service.The common path planning problem is to find an optimal or near-optimal path from the initial state to the target state to avoid obstacles in the working environment with obstacles according to some optimization criteria(such as the minimum energy consumption,the shortest journey,short task time,etc.).At present,the global path planning algorithm includes visual method,grid method,neural network aIgorithm,sequential quadratic programming method,etc.These methods have their own advantages,but they all have some limitations,which are easy to fall into local optimal.Genetic algorithm(GA)is a method to search for optimal solution by simulating natural selection and biological evolution.When solving complex combinatorial optimization problems,it has the advantages of robustness,flexibility and not easy to fall into local optimum.Although genetic algorithm using crossover and mutation operator can search the optimal solution from the global perspective,but genetic algorithm time and number of chromosomes genetic operator such as exponential growth,the relationship between the choice of chromosome encoding and largely limits the solving efficiency of genetic algorithm,and poor local search ability,easy generation "super individual" form a precocious phenomena.In this paper,a hybrid genetic algorithm is constructed,which takes both global and local perspectives into account.The global search ability of genetic algorithm and the strong local search ability of simulated annealing algorithm are combined to design the algorithm,so that the individual searched is closer to the optimal solution.At the same time,it can improve the solving efficiency of genetic algorithm and avoid the problem of precocity of genetic algorithm and slow search speed of simulated annealing algorithm.This paper designs two application scenarios to verify the effectiveness and superiority of the hybrid algorithm.One is in a small satellites flying off the space target environment,using the hybrid algorithm design optimization of spacecraft orbit,the security,the path of the spacecraft rail task time,fuel consumption,total distance constraint conditions,such as planning the conform to the constraints of spacecraft orbital path,and the efficiency of the proposed hybrid algorithm is superior to the literatures of the standard genetic algorithm;Two is considered under the influence of the disturbing force,service spacecraft without implementation orbital maneuver,to visit close to a space target,and make use of the hybrid algorithm to solve optimal perturbation visit close to orbit,the simulation results show that the proposed hybrid algorithm can not only in the global search to the optimal perturbation approach orbit,and can be in local optimization more optimal solution,and find out the service of the spacecraft without implementation orbital maneuver multi-objective optimal perturbation close to visit space orbit.