Research On The Theory And Technology Of Satellite Cooperative Localizaiton In Space-based Sensor Network

In the Space-based Sensor Network for the purpose of maintaining uniform time and space system,the maintenance of the space-time system in the satellite subsystem plays an important and fundamental role.Exploiting inter-satellite precision ranging to implement Cooperative Localization(CL)over satellites can make the entire constellation maintain a usable space-time reference for a long time without ground support,which has high practical significance and strategic value.Due to the nonlinearity of the satellite motion model,the traditional CL theory and analysis methods cannot be directly used to analyze the performance and error evolution characteristics of CL in satellites,which restricts the development of CL in satellite constellation.In addition,in order to achieve Satellite Cooperative Localization(SCL)in practice,it is necessary to design a high-performance distributed cooperative algorithm,which makes the algorithm and system more flexible while providing sufficient positioning accuracy.Furthermore,since the link and power resources are very limited in satellites,it is also a very important issue to optimize to reduce resource consumption as much as possible while keeping the performance level.According to the above analysis,this thesis focuses on the following key issues:1.The theoretical analysis framework of the observability and error evolution of SCL: Theoretical analysis can demonstrate the feasibility of SCL and provide with an analytical way for studing the error evolution of SCL.However,due to the nonlinear characteristics of satellite motion model,the traditional CL framework cannot be directly applied to the analysis of SCL.In addtion,the existing analysis methods for SCL can only be used in the state space of Kepler orbital elements for some special cases,and it is difficult to analyze the error characteristics of SCL analytically.Therefore,analytically studing the observability and error evolution of SCL in dynamic situation is the first key issue in this thesis.2.The design of distributed filtering algorithm for SCL: In the engineering practice of SCL,distributed algorithms have better invulnerability and flexibility than centralized algorithms.However,the performance of traditional distributed SCL algorithms have their own disadvantages.On the one hand,average distributed algorithm can not reach the same performance level as the centralized algorithm.On the other hand,distributed algorithms with similar performance as centralized one place a high requirement on the communication resources of the inter-satellite link,which are very limited.To this end,how to design a new type of distributed SCL algorithm to further reduce the consumption of inter-satellite communication resources under the premise of improving positioning accuracy has become the second key issue in this thesis.3.Inter-satellite link optimization for SCL: The measurement information that different inter-satellite links can provide is different.Under the same conditions,the intersatellite links between satellites that are far apart can provide less ranging information than the satellites that are closer,i.e.,the ranging error is larger.Different configurations between linked satellites can also affect the performance of satellite positioning due to the directionality of the measurement information.The position error of the satellite is also included in the inter-satellite measurements.These factors have brought necessity and complexity to how to optimize the selection of inter-satellite links.How to choose the appropriate metrices and optimize the inter-satellite links accordingly becomes the third key issue in this thesis.For the theoretical framework,this thesis firstly proposes a theoretical analysis framework for SCL performance and error evolution analysis in Cartesian coordinate system,which can analyze the SCL performance and the nature of error evolution in an analytical way.As for the filter design,a distributed algorithm based on state caching is proposed,which can obtain higher positioning accuracy than traditional distributed methods under the condition of using only few link communication resources.For the optimization of the inter-satellite links,a series of metrices for optimizing inter-satellite links are also proposed,and heuristic and convex optimization-based methods are proposed to optimize these metrices.These technical methods can solve the corresponding three key technical issues above.These propositions have been verified through simulations,which proves their technical feasibility and effectiveness,and can provide theoretical guidance and design for SCL and space-based sensor network.