Coordinated Task Planning Research Of Space-aeronautics Earth-Observing

Space-aeronautics Earth-Observing technique is an important means of perceiving the real world directly, which has been rapidly developed in recent years. With a growing diversity of the platform and the load, the maneuvering capability, onboard processing capability and data-transfer capacity are enhanced significantly. Moreover, the observation precision, width of the load and aerospace information acquisition ability are greatly improved. The new trend of earth-observing must be multi-platform and multi-sensor coordinated earth observing. Several factors, including the user requirements, the resources capabilities and the various resources constraints must be considered. On this basis, the observation plan should make full use of complementary advantages of multi platforms, and maximize the value of the observation timeliness, the spatial resolution, the temporal resolution, the spatial coverage and the spectral coverage. How to make this plan is a crucial and real problem. So the space-aeronautics resources coordinated planning method has been an important research part in the modern earth observing technique field in recent years. Therefore, it is important actual significance and practical application to study of the space-aeronautics resources coordinated planning method.Space-aeronautics cooperation task planning method for earth observing aims to improve the efficiency and quality of observation tasks and maximize the value of space-aeronautics resources through allocating space-aeronautics resource in reason, and fully takes into account the user requirements. Based on the coordination theory, the modeling theory and optimization theory, represented by satellites and UAVs, the paper mainly focuses on space-aeronautics cooperation task planning method of the multi-phase observation task and emergency observation tasks. This thesis contains the following works and innovations:(1) The model of space-aeronautics cooperation task planning problem for earth observing is established.After the earth-observing performance comparison between satellites and UAVs, the space-aeronautics cooperation task planning problem for earth observing is posed. Based on the hierarchy analysis of space-aeronautics cooperation earth observing system, the system can be seen as a Holonic organization model. The Holonic organization model is analyzed in details from three perspectives: Top-Down, Horizaontal and Buttom-Up. And then three key problems for cooperation task planning are concluded: cooperation problem between father-holons and sub-holons, cooperation problem between sub-holons and sub-holons, and cooperation problem within sub-holon systems. Considering the actual situation of space-aeronautics cooperation earth observing system, the traditional horizontal cooperation task planning framework 第 5 页 is improved for multi-phase observation tasks. Besides, the user-oriented vertical cooperation task planning framework is proposed for emergency observation tasks. The key problem elements are analyzed in details, including satellites resources, UAVs resources, observation tasks and optimization goals. On this basis, the multi-phase observation task planning model and emergency observation task planning model are established. And the input information, decision variable, constraint condition and objective function of the model are specified and mathematical formalized. This model could serves as a foundation for the further research.(2) A satellite schedulability prediction algorithm based on the robust decision tree and Bagging Support Vector Machine(Bagging SVM) is proposed.With notably few exceptions, the existing satellite mission operations cannot provide the ability of schedulability prediction. The coordinator can do nothing but waiting for the results of time consuming batch scheduling. Satellite task scheduling typically takes several hours to schedule one day‘s activities for a set of satellites and tasks. It is often too late to adjust the request when receiving scheduling failures. Therefore, based on the historical planning information, how to predict a given request‘s likelihood of being included in a near-term schedule is an urgent problem of Earth Observation satellite task planning research.To solve this problem, a supervised learning algorithm based on robust decision tree and Bagging Support Vector Machine(Bagging SVM) is proposed. The Bagging SVM is applied to improve the accuracy of classification, and robust decision tree is utilized to reduce the error mean and error variation. The simulations and analysis show that a prediction action can be accomplished in near real-time with high accuracy. This means it can provide the prediction information for coordinated task planning, and the decision makers can maximize the probability of successful scheduling through changing request parameters or take action to accommodate the scheduling failures in time.(3) The heterogeneous MAS corporation task planning algorithm based on the market model is proposed for multi-phase tasks. On this basis, the space-aeronautic coorporation earth observing task planning algorithm is proposed.Coordinated observation of air and space assets is the trend of earth observation and it is expected to continue in the future. In order to increase the information gain of earth observation and improve the completion ratio of multi-phase missions, this paper analyzes the heterogeneity of observation resources and the diversity of decompositions for complex observation missions. Considering the differences between a satellite task planning model and a UAVs task planning model, the problem can be modeled as a heterogeneous MAS multi-phase cooperative planning problem.Based on the analysis above, the characteristics of heterogeneous MAS multi-phase cooperation task planning problem is analyzed, and the problem model based on the Markov decision process theory is proposed. Subsequently, the optimization object of the cooperative planning system is analyzed, and the negotiating relationships of the different objects in planning process are discussed. On this basis, a heterogeneous MAS multi-phase cooperative planning algorithm based on the market model is proposed. Nevertheless, with the increasing number of sensors and observation requests, complexity of the problem increases. Moreover, the optimization model and the constraints of different platforms are varied. Centralized optimization algorithms sometimes are powerless. To tackle these difficulties, we explore the iteration coordination mode of algorithm according to the Communicating Sequential Processes(CSP) method, and propose an adaptive ―Super Step‖ iteration coordinate mode. The stability of proposed iteration coordinate mode is proved based on the CSP theory.On this basis, the space-aeronautic coorporation earth observing task planning algorithm framework based on the HMAS-MPCPP is proposed. Besides, the local task planning algorithm is given according to the characteristics of satellites and UAVs, and a complete space-aeronautic coorporation earth observing task planning algorithm for multi-phase observation task is accomplished. Finally, the above methods are used to solve the joint observation problem of air and space assets. Experiment and analysis show that the proposed approach can solve the problem effectively.(4) An algorithm framework is presented using label- constraint shortest path technique with the conflict resolution. On this basis, the space-aeronautic data acquisition chain task planning algorithm for emergence observation tasks is proposed.Firstly, the observation data acquisition process of satellite and UAVs is analyzed in detail, and the space-aeronautic data acquisition chain task planning model is established. Considering there are many approaches for data acquisition, the space-aeronautic data acquisition chain task planning problem can be seen as a multi-modal route planning problem. Then, the algorithm for label- constraint shortest path problem with the conflict resolution(LCSP-CR) is proposed. Based on the framework of this algorithm, considering the characteristics of the satellites and UAVs, through a series of implementations and optimizations to the initial LCSP-CR algorithm framework, the space-aeronautic data acquisition chain task planning algorithm is proposed. The exact solution can be obtained in polynomial time using the proposed algorithm. Experimental results proved the powerful performance of proposed algorithm in which the demanding of space-aeronautic data acquisition chain task planning could be done in less than 1 minute. These technologies above is the foundation for the practical application in the future.