Research Of Space-air Remote Sensing Resource Intelligent Mission Planning Approach

In recent years,the requirements of remote sensing data increase rapidly.Alongside with the increasing requirements,the timeliness condition of requirements,comprehensiveness of information and complexity of mission planning increase as well.Multi remote sensing resources are needed to achieve high-level remote sensing requirements in highly cooperative circumstances,especially in land surveying,emergency rescue and disaster prevention domains.At the same time,the clients of remote sensing services become more and more unprofessional.As a result,the remote sensing resource control system needs to have the abilities of autonomous analyzing of user requirements,mission planning and resource scheduling with low computation costs.Based on these requirements,in this thesis,the intelligent mission planning and scheduling approach for space-air based cooperative remote sensing missions is proposed.The aims of this research are helping remote sensing resource control system to analyze user’s requirements,planning the remote sensing mission and scheduling remote sensing resources based on the current system conditions with low computation costs.This approach also has the ability of intelligent optimizing the remote sensing plan based on characteristics of different resources and mission scenarios.To achieve these goals,a comprehensive background research is first conducted.Then,this problem is divided into four phases: space-air resource and mission selection phase,space-air mission assignment phase,space resource mission planning phase,and air resource mission planning phase.In space-air resource and mission selection phase,due to customer requirements,the space-air resources and remote sensing missions are selected.In space-air mission assignment phase,remote sensing missions are assigned to different kinds of space-air resources.In space resource mission planning phase,due to the characteristics of space remote sensing resources,the mission plan of space remote sensing resources is optimized.And in air resource mission planning phase,the control system of air resources is optimized to execute remote sensing missions using a simulationbased evolutionary algorithm.The main contributions of this thesis include:a)To solve the problem in space-air resource and mission selection phase,deeplearning-based space-air resource allocation method is proposed.This method has good algorithm performances and low computation costs.In this method,the space-air resource allocation problem is first transformed into a vector using Graph Neural Networks.Then,the allocation plan is made using a deep-learning-based method.At last,Simulation experiments are designed,and experiment results are reported.b)To solve the problem in space-air mission assignment phase,an ant-based multi objective heuristic algorithm is implemented.First,the problem characteristics of spaceair mission assignment problem are analyzed.Then,a Periodic Vehicle Routing Problem based problem model is illustrated.To solve this model,an improved Ant Colony Optimization algorithm combined with Multi-Objective Simulated Annealing is proposed.At last,Simulation experiments are designed,and experiment results are reported.c)To solve the problem in space resource mission planning phase,a Multi Ant System based heuristic algorithm is implemented.First,the problem characteristics of space resource mission planning problem are analyzed.Then,the problem is modeled as a special variant of Vehicle Routing Problem with Time Windows.A Multi Ant System based hybrid heuristic algorithm is proposed to solve this model.At last,Simulation experiments are designed,and experiment results are reported.d)To solve the problem in air resource mission planning phase,a rule-based air resource control model is implemented and a multi objective heuristic algorithm is applied to improve the performance of control model.First,the problem characteristics of air resource mission planning problem are analyzed.Then,a flexible and powerful air resource control model is illustrated.A multi objective optimization heuristic algorithm is applied to tune the control parameters of this model.Simulation experiments are designed to evaluate the performances of the optimization heuristics and controls model.