Research On Mission Planning Modeling And Solving Method Of Agile Satellite For Regional Intensive Task

Mission planning plays an important role in the observation control system of an agile satellite,whose quality directly affects the observation performance of an agile satellite,especially in the regional scenario of intensive tasks,such as disaster prevention and reduction,region combat,counter-terrorism and stability,and joint operations.Limited by the hierarchical structure of existing mission planning models,the information delivery between mission scheduling and attitude planning is unidirectional.The indirect information delivery makes the mission scheduling unable to use the information form attitude planning when searching an observation scheme,which results in the low information utilization rate and low mission planning efficiency.Attitude changing time is a decisive judgment for the observation scheme of mission planning,and its accuracy and precision directly affect the determination of observation schemes.The existing models of attitude changing time have some limitations,such as parameter dependency,unlimited torque,uncoupled calcualtion.All of these limitations cloud result in torque overflow and satellite operation failure.Aiming at overcoming the above problems,this paper focuses on the modeling and solving method of agile satellite mission planning problem for regional intensive task and attitude changing duration sub-problem,and verifies the effectiveness of the proposed models and methods through typical scenarios.The main research contents are as follows:(1)The attitude coordinate system,kinematic model and agile satellite working principle are summarized,the main mission planning elements of agile satellite for regional intensive targets are then explored,the mission planning work-flow is designed finally.Based on the definition of the coordinate system,establishment of attitude motion model and introduction of satellite working principle,this part elaborates the mission planning from three aspects,objectives,constraints and input-output,and gives the basic assumptions and simplifications.On these bases,the mission planning work-flow for regional intensive targets is designed.(2)The classic models of attitude changing time is discussed and analyzed.Aiming at overcoming their shortcomings including parameter dependency,unlimited torque,uncoupled calculation,a Bolza model is proposed,and its solving method based on pseudospectral is designed.Firstly,the limitations of the existing models are summarized in terms of attitude changing parameter dependency,unlimited torque,uncoupled calculation.Then to overcome these shortcomings,the Bolza model and two-stage Bolza model of the attitude changing time are respectively proposed for two cases of overlapping time intervals and non-overlapping time intervals.Finally,a pseudospectral method for the Bolza models is designed,and simulation tests verify that the proposed model and method can get rid of the dependence of the parameters.(3)To overcome the shortcoming of low information utilization ability of the hierarchical planning structural,a TSP-based collaborative mission planning model is proposed for the intensive point tasks under staring mode,and then a genetic-pseudospectral method is designed to solve it.Firstly,the mapping relationship between the studied problem and TSP is established,based on this a TSP-based intensive point tasks mission planning model is proposed with the aim of minimizing energy.Considering the collaborative structure of the model and the advantages of the attitude kinematics method in obtaining the energy-optimal attitude path,a hybrid method named genetic-pseudospectral method is proposed.In this method,the permutation-based two-dimensional coding structure is designed to provide scheduling scheme and observation time,while a pseudospectral method is used to provide attitude execution and energy consumption.In addition,a chromosome feasibility analysis and illegal chromosome repair principle are constructed to avoid the phenomenon that all solutions in a generation are in-feasible,and a penalty function is introduced to ensure that genetic search can be performed in multiple directions.Simulation results show that the proposed algorithm is superior to the classic genetic algorithm in terms of energy consumption,attitude path smoothness,and observation number of tasks.(4)A further consideration for the regional intensive stripe tasks under under scanning mode is discussed.To enhance the information utilization capacity of the mission planning model,a mission-attitude mapping based intensive stripe tasks mission planning constraint satisfction model is proposed,and then a preference-based differential game evolution algorithm is designed to calculate it.Firstly,a time-varying attitude path is designed to express the stripe task,based on which the constraint satisfaction model is used to establish intensive stripe tasks mission planning model.And then,a preference-based differential game evolution algorithm is proposed.In this algorithm,a series of differential evolution operators is designed to satisfy the permutation of decision variables,while the game theory is introduced to establish a multi-objective compromise utility function with an evaluation mechanism.Meanwhile,a preference-based distance evaluation mechanism is designed for individuals with the same fitness,which can keep farthest individual in the population.Simulation results show that the proposed algorithm can better obtain the win-win solution for multiple optimization goals compared with the simple differential evolution more efficiently.