Research On Loop Sum-product Cooperative Navigation And Positioning Method Of Message Transmission

Compared with a single-platform system,a multi-motion platform system can complete more complex tasks and has higher performance.Cooperative navigation is a navigation method suitable for multi-platform systems.The measurement information is fused to obtain the best estimate of the entire system.Traditional collaborative navigation uses a centralized or hierarchical architecture,and a single central node is used to process the data of the entire system.Although the structure is simple,it requires high computing capabilities for the central node.The entire system is extremely dependent on the central node and is only applicable for small systems with a small number of platforms.The distributed collaborative navigation does not have a central node,and each internal platform is a processing node.Compared with centralized or hierarchical collaborative navigation,it has higher stability and robustness and can be applied to systems with a larger number of platforms.The disadvantage is that it requires high real-time communication capabilities of the system,and wireless communication often consumes energy,which also has a certain impact on the endurance of the system.Once the amount of communication between platforms is reduced,the navigation accuracy may decrease due to the decrease in observation information.The difficulty of decentralized collaborative navigation algorithms is to strike a balance between communication volume and navigation accuracy.This paper designs a multi-platform distributed collaborative navigation scheme in three parts.The first part is absolute navigation,which uses the platform's sensors to perform preliminary positioning and obtain a preliminary understanding of the entire distributed system;the second part is relative navigation,using the inter-platform.The sensors measure the relative motion information between the platforms.In this step,all platforms are connected in series to get a preliminary optimization of the internal relationship of the entire system;passing,on the base of the information obtained in the first two steps,the final step is message passing algorithms for navigation and location.To reducing the flow of message delivery as much as possible,the amount of calculation is allocated on each platform.And the local optimum solved by a single platform is used to approximate the entire system's global optimum distribution.In absolute navigation,firstly,inertial navigation and satellite navigation are modeled.Error analyzed,combined with the two's advantages,a tightly combined state-space model is established,and the pseudo-range rate information of the satellite's original output is used to pass the untracked Kalman Filtering is used to estimate the error of inertial navigation and make feedback correction,which obtains the accuracy of inertial navigation but also has the stability of the satellite system.In relative navigation,the relationship between the pseudo-range double-difference made by the dual-machine dual-star and the spatial relative vector of the two platforms,coupled with the high-precision ranging of the ultra-wideband technology,is used to model the relative motion state between the two platforms.Combine the unscented Kalman filter algorithm to solve and estimate the relative state.Given the possible delay and packet loss problems in the measurement process,the Bernoulli sequence is used to model the actual measurement.When the measurement does not arrive on time,the last time measurement is used,and the one-step delay of the measurement is used.Prediction replaces the measurement at the current moment for the solution.In collaborative navigation,use the results of absolute navigation and relative navigation estimation to establish a state-space model and transform it into a factor graph model.Based on the loop and product algorithm,message passing is carried out through the communication link between platforms,thereby estimating the overall system Optimal distribution.Finally,the algorithms and effects of each part are verified through the simulation of the MATLAB platform.The simulation data prove the superiority and feasibility of the collaborative navigation scheme proposed in this paper.Compared with the traditional navigation and positioning system and the cooperative navigation scheme,it has higher anti-interference performance and stability and can adapt to different complex environments.