Flight Sequence Planning Method For Large-scale-object Visiting Mission

Own to the high exploration efficiency and low average cost,one flight to visit multiple targets is increasingly polular and will become the main type to implement the on-board servicing and deep space exploration missions in the future.The design of multi-target visiting mission,especially the large-scale-target visiting mission is a very challenging work,where the planning for the visiting sequence is the core.In order to improve the design ability for complex space target visiting mission,the flight sequence planning method for large-scale-object visiting mission is studied in this paper.The main research contents and results are listed as follows:A method for estimating the optimial velocity increments of perturbed multiple-impulse rendezvous based on multiple neural networks is proposed.First,the relationship between the optimal velocity increments of perturbed multiple-impulse rendezvous and the initial right ascension of the ascending node(RAAN)difference between the departure body and the rendezvous target,as well as the transfer time,are investigated and revealed.It is first found that perturbed multiple-impulse rendezvous can be divided into three types,which are “RAAN-closing rendezvous”,“RAAN-intersecting rendezvous” and “RAAN-separating rendezvous”.Then,the method for estimating the optimial velocity increments using "respective learning +overall estimation" strategy is proposed.An efficient database generation method to obtain three types of optimal solutions is developed,and the estimating flow based on three regression multi-layer perceptrons is presented.The simulation results show that the estimation accuracy of the proposed method is much higher than analytical estimation methods.The estimation error of single rendezvous can be reduced to lower than 3%,and the estimation error of rendezvous sequence is only 0.3%.A step-by-step planning method for multi-target rendezvous sequence based on a dynamic sequence planning ant colony optimization algorithm(ACO)is proposed.First,an improved ACO for solving moving-target sequence planning problem is designed based on the framework of ant conlny system.A pheromone tensor for characterizing the transfer preference between each two target in dynamic sitiation is employed,and the corresponding solution construction and pheromone updating method are designed based on the pheromone tensor.Then,a step-by-step rendezvous sequence planning strategy based on the time discretization is proposed,and the overall multiple-impulse multi-target rendezvous sequence planning flow is presented.The simulation results show that the proposed method owns stronger optimization ability than mix-integer genetic algorithm for complex rendezvous sequence planning problem.A group-by-group planning method for multi-group rendezvous problem method based on a packing programming ACO is proposed.First,a packing programming ACO for solving multi-group rendezvous mission planning problem is designed based on the framework of max-min ant system.A triangular matrix for characterizing the preference degree of putting two targets into the same group is employed,a self-adaption mechanism for dynamically adjusting the algorithm parameters is added,and the optimization performance of the algorhithm is improved.Then,a step-by-step multi-group rendezvous sequence planning flow based on the moving-target sequence planning ACO is presented,and a group-by-group planning strategy that can gradually improve the solution quality based on the four-step planning flow is designed.A nearly global optimization of the grouping scheme and rendezvous sequence in each group for large-scale visiting targets is realized.The proposed method helps the author get the second prize in the ninth global trajectory optimization competition.An accumulated search method for multi-target flyby sequence based on the transformation of rendezvous trajectory optimization is proposed.First,a multi-target flyby trajectory planning strategy based on the transformation of rendezvous trajectory optimization is proposed.The trajectory planning methods for one impulse to fly by two targets based on a four-impulse two-target rendezvous model and two impulses to fly by three targets based on a seven-impulse three-target rendezvous model are designed.Then,a multi-target flyby segment searching method based on“ shooting+adjusting” is designed,and a multiple-impulse multi-target flyby sequence searching method based on the accumulation of the flyby segments is proposed.A nearly global search for the flyby sequence containing one-impulse multi-target segment is realized.The proposed method helps the author win the first prize in the tenth global trajectory optimization competition.Although the research is guided by typical problems in this paper,the proposed methods are universal and flexible in some degree.The research results can provide strong technical support for our country to implement more complex space multi-target visiting mission in the future.