Research On Autonomous Task Planning Of Small Satellite Based On The Improved Genetic Algorithm

In recent years,with the surge in the number of small satellites,ground stations' load of monitoring and controlling small satellites has been growing,so small satellites need have the ability of autonomous operation.Onboard autonomous task planning is the foundation of autonomous operation.Satellite task planning has been proved to be a kind of multi-constrained time-related NP-hard problem.Intelligent planning algorithm is an effective way to solve this kind of problem.In this paper,we focus on the task planning modeling,the individual coding and decoding rules,the crossover and mutation strategy,and the resource allocation in long-period task planning,to solve the problem of small satellite autonomous task planning,considering time window constraints and dynamic resource constraints.The following research results have been obtained:Considering the complex diversity and temporal relevance of satellite tasks,a coding rule based on fixed-length integer sequence coding is proposed.The rule uses the fixedlength integer sequence to encode the planning task sequence,compared with the floatingpoint code,Integer encoding and decoding rules narrow the search space,due to the finiteness of full permutation of one integer set.The local genetic information transfer characteristics,post dominant decoding are analyzed,which provides a theoretical basis for the improved genetic algorithm.Aiming at the requirement of satellite on-orbit application,a improved genetic algorithm with multi-mode crossover and mutation is proposed.The algorithm classifies individuals according to the satisfaction of time constraints and resource constraints.And crossover and mutation operators are designed for each type of individuals.Because of the partial sorting and individual classification,we can exclude some bad individuals in advance,narrow the search space and improve the efficiency of the algorithm.Based on the adjustable group roulette and elite retention strategy,a selection strategy is designed.Because of the compromise between the traditional roulette operator and the random selection operator,the genetic algorithm can improve the early maturation problem.Aiming at the continuous planning and resource allocation of satellite tasks,based on the improved genetic algorithm,a framework of task planning algorithm based on rolling planning is designed,and a rolling planning algorithm,a resource preallocation algorithm and a resource redistribution algorithm are developed.And the experimental results show that the feasibility of the proposed framework and the effectiveness of the proposed algorithm are shown in this paper.