Model And Algorithm Research On Earth Observing Satellite Mission Planning

The mission of planning and scheduling with earth observing satellite is to allocate the resources including satellite and ground station optimally according to the users'requirement, so as to satisfy multifarious requirements in using satellite in the future. At present, multi-mission planning is a noble research domestically and overseas. With this research, planning and scheduling with multiple satellites could be studied theoretically and the practical requirement in using satellite could also be satisfied. Conventional planning with satellite focuses on definite missions and targets and achieved little in studies of planning of unknown moving targets which were studied here in this research. The main contents and creative results are outlined as following.Firstly, based on introduction and analysis of general planning techniques and search theory, mission planning with satellite was elicited and divided into two main steps: pretreatment and optimization. The task of pretreatment is to screen satellites resources according to user's requirement and to determine the available resources and time slots for each mission. Whereas, the objective of optimization is to determine which mission is to be scheduled and to allocate resources and time slots for the scheduled mission.Secondly, three models were established to describe the simplified questions during the study of optimization, including mixed integer programming model, constraint satisfaction optimization programming model and a combining new model. Distinct solution algorithms were worked out for different models. Modified Lagrangian relaxation algorithm was used for mixed integer programming model. Simulated annealing algorithm was used for constraint satisfaction optimization programming model. Heuristic algorithm was used for the new model.With regard to planning of moving targets, moving targets in the sea were set as the main object. Planning of moving targets in the sea was studied and worked out through the following four steps. The motor region was determined by analyzing the target's movement. Then probability distribution of the target in the region was determined. The region was partitioned by considering the coverage of satellite. And then search strategy was deployed for solution. Concurrently, a method of calculating detection function was given by combining entropy of information and membership function.Finally, the intact process multi-mission planning with satellite was clarified through a designed simulation example. The planning models and algorithms developed in this study were proved to be feasible and effective with plenty of data and other simulation validation.