Study On Modeling And Solving Approaches For Planning Of Space Station Onboard Missions

The planning of onboard missions is a key technology to ensure the normality of the space station’s long-duration operation.With the background of China’s space station project,this dissertation studies the planning problems of space station onboard missions,aiming at the short-term planning requirement.It mainly focuses on the planning domain modeling of onboard missions,the timeline heuristic planning,the evolutionary optimization of onboard missions,and the mission robust planning.The main achievements obtained in this dissertation are summarized as follows.Based on the ontology theory,the planning domain modeling approaches of space station onboard missions are proposed.At the beginning,the major elements of the planning domain,such as the mission,process and action,astronaut and device,resource,constraint,objective,algorithm and period,are analyzed.Further the characteristics of the timeline,mission,resource,device,constraint,and planning of the domain are summarized,and the Knapsack mathematical model of the planning problem is built.Finally,based on the ontology theory,an ontology planning domain model of the space station onboard missions is developed.Based on the characteristics of concepts,the object concept ontology model and method concept ontology model are represented.The XML description of the domain shows that the proposed models can express well the planning domain knowledge of space station onboard missions.Based on the timeline,the classical heuristic planning approaches of onboard missions are proposed.The mission states in the timeline are analyzed,and the time redundancy heuristic strategy,arranging the missions as early as possible,is defined.Meanwhile,the plan iteration way and the new plan generation way are proposed to solve the problem.Based on the plan iteration way,the time iterative planning model for conflict-resolving is developed.Based on the proposed initial plan and iterative conflict-resolving strategies,the problem is solved based on the initial plan according to the time iterative process.The results show that the proposed approach can obtain the solution satisfying the constraints efficiently and fast.Based on the new plan generation way,the greedy planning model of onboard missions is built,the time-relaxing heuristic strategy is defined,and the constraint propagation model is presented.Finally,the planning problem is solved separately by the greedy method using the time-relaxing heuristic strategy and by the greedy method using the priority strategy.The results show that the variations in constraints in the planning process can be analyzed efficiently by the constraint propagation model,and the greedy methods based on the time-relaxing heuristic strategy and based on the priority strategy can both obtain the feasible solutions satisfying the heuristic requirements.Based on the evolutionary algorithms,the single-objective and multi-objective planning approaches of onboard missions are proposed.The optimization contents in the planning domain of onboard missions are analyzed.For the multi-constraint feature,the constraint satisfaction model is built,and the backtracking method is used to meet the constraint satisfaction requirement according to the defined plan states and the presented backtracking conflict-resolving strategies.After that,the single-objective optimization model is built,in which the objective value is the sum of missions’ priorities in a plan,and a single-objective global optimization way with the backtracking method and genetic algorithm is used to optimize the planning problem.The results show that the backtracking method can obtain a solution satisfying the constraints,while the single-objective optimization way can obtain a solution with better optimization performance.Furthermore,the multi-objective optimization model is built,in which three objectives,including the Priority index,Profit index,and Stability index,are considered,and a multi-objective optimization global way with the backtracking method and Non-dominated Sorting Genetic Algorithm II(NSGA-II)is adopted to obtain the multi-objective optimal solutions,and the results successfully reveal the trade-off relationships among the objectives.Considering the perturbation of astronauts and devices,the robust planning approaches of onboard missions are proposed.The executing delay problem of onboard missions is analyzed,the concept of perturbation delay is defined,the assumptions for robust planning,including the ―Immediate Handling‖ and ―Resolving Accomplish‖,are presented,and the robustness of plan execution is analyzed.After that,the generation model of mission delay perturbation is built based on Poisson distribution,the delay resolving time model is built based on Exponential distribution,based on them,the Complete Execution Probability robust index is defined.The two-objective optimization method is used to optimize the Priority index and Complete Execution Probability index,and the Complete Execution Probability indexes of different methods are compared.The results show that the delay perturbation research of onboard missions can obtain the feasible solutions considering both the Priority index and Complete Execution Probability index requirements,and the effectiveness of the proposed robust planning approach is validated.Furthermore,the delay perturbation influence for the constraints are analyzed,and based on the delay strategy of precedence relationship,the onboard mission robust planning model considering the constraint propagation is built.The results show that the impact of the delay perturbation on the change of constraints is obvious,and the great influence of astronaut onboard working ability on the Complete Execution Probability of the plan is revealed by comparison tests.Based on the engineering requirements of China’s space station program,this dissertation has studied the modelling and planning approaches of onboard missions for the short-term planning requirement,and will be helpful for the operation management of China’s future space station.Thus,the achievements obtained have important engineering application values.This dissertation summarizes the planning problem essence of onboard missions from the space station operational engineering practice,abstracts numerical planning models,proposes efficient planning approaches,and therefore,has some theoretical significance.