Research On Distributed Satellite System Formation Adjusting Planning

Utilizing given formation to accomplish mission goal is a representative characteristics of Distributed Satellite System (DSS). DSS takes good technical advantage of modern micro satellites, extends their applying area, increases mission reliability and damage resistibility, and decreases mission cost effectively, while these advantages of DSS are tightly correlated to proper formation and formation adjusting ability, so formation adjusting is one of the key technique of DSS. Formation of DSS includes position formation and attitude formation, and formation adjusting includes formation initialization, formation keeping, formation reconfiguration and formation maneuver. In this thesis, systematic research on formation adjusting planning was conduct, main contents and accomplishments are as follows:1. Autonomous formation keeping planning method based on multi agent negotiation was proposed. Each satellite was mapped to an Agent, which autonomously percept its relative motion state and planned its own adjusting strategy, meanwhile put forward the effect, cost and inclination for its optimal adjustment. Then negotiation was conducted based on these criteria and given rules, such the adjusting strategy was selected. Simulation indicates that this mechanism of autonomous distributed planning and coordination accomplishes relative formation keeping task well, and multi agent negotiation provides a promising way for DSS autonomous formation planning.2. Multi-impulse adjusting problem of DSS was systematically investigated from single satellite adjusting to multiple satellite adjusting. For single satellite adjusting, expression and handling of the complex expected formation constraint were discussed. Advantages and disadvantages of different optimization methods and constraint handling methods were analyzed through examples. After that, an integrated way of combining to global search ability of evolutionary algorithm and local search ability of traditional method was proposed. For complex multiple satellite adjusting, indexes of multiple satellite co-ordination were mainly discussed, including fuel inequality, target location assignment, collision avoidance and time constraint. Solving example problems, the integrated methods get excellent results in both global optimal and local constraint satisfaction, which shows the effectiveness of proposed indexes and methods.3. Distributed collaborate planning method for DSS formation adjusting was put forward. Formation adjusting problem was divided into two layers planning, i.e. collaborating top planning with satellite location assignment and first satellite phase optimization, and bottom planning with single satellite fuel optimal adjusting. The coordination of multiple satellite bottom planning and the coupling of top and bottom planning were represented by response surface models (RSM). Top planning was conducted based on RSM built by bottom planning, then more samplings were computed by bottom planning around solution of top planning to built more accurate RSM, and new top planning was conducted. That iterative planning searches valuable region with high priority, thus saves computation greatly meanwhile ensures solution precision. Examples show that the method effectively reduces problem solving complexity and computing cost, and evidently improves searching result. By parallel execution of single satellite sub optimization, more computing time will be saved. This provides a valuable way for autonomous on-board planning.4. Methods of Analytical Mechanics were used in the relative motion of DSS. Hamilton modeling and generating function method for relative motion were introduced. Approximation of Hamilton function for relative motion of satellite formation and its impact on modeling accuracy were interviewed. Advantages include that these methods adapt to elliptical orbit well, and take high order perturbations naturally, which are bottle necks of traditional methods of Newton Mechanics (such as, Hill's equations). By two adjusting problem of circular reference orbit and elliptic reference orbit respectively, effectiveness of the modeling and solving method was testified.5. Optimal control of DSS formation adjusting based on Hamilton Mechanics and generating function method was investigated. An iterative way of solving the two-point boundary value problem induced by optimal control by relative motion generating function approximation was proposed. First low order approximation of generating function and approximate optimal trajectory were obtained. Then using low order approximation of generating function of Hamilton "relative motion" to the approximate optimal trajectory, improvement of the approximate optimal trajectory was iteratively made. In this way, high order approximation accuracy can be obtained by only using low order approximation, which is much cost effective in computing. Solving of a circular reference orbit problem and an elliptical reference orbit problem both reveals that high control precision is reached.6. Aiming at the need of DSS simulation, a simulation software framework was designed. It adopt the data distributing service and the publish-subscribe mode. Cooperativity and reusability of both inter-process and inner-process simulation member were considered. So that an environment can be developed uniformly supporting DSS operation simulation, which contains both strong and weak coupling models, and has improving real time simulation requirements along with developing process. Based on the simulation software framework, a simulation system of DSS dynamics and control were partly implemented. And effectiveness of proposed high precision optimal control of DSS formation was testified.