Reserach And Implementation Of Adaptive Integrated Navigation System With IMU-based Auxiliary Measurements Detection

Nowadays,INS/GNSS integrated navigation system is one of the widely used navigation solutions.It overcomes the limitations of the navigation system based on one single sensor which is difficult to conduct long-term and high-frequency navigation.This integrated system improves the reliability of navigation information,shortens the information acquisition time while reducing the navigation cost,and improves the navigation accuracy to a certain extent.However,once the GNSS signal outage,the system will switch to inertial navigation mode again.This makes the navigation accuracy of the current integrated system diverge rapidly due to its poor stability.Therefore,a robust adaptive integrated navigation system based on low-cost sensors is developed for various navigation scenarios.This dissertation proposes an auxiliary update detection algorithm based on IMU.This algorithm uses neural networks to detect the static and turning behavior of vehicles based on IMU measurement data,and then converts them into zero-velocity measurement and non-holonomic constraints measurement,separately.In order to solve the limitation of the traditional integrated navigation system,the measurement is combined with Kalman filter.Supported by the above algorithm,this dissertation designs and implements an adaptive integrated navigation system,which is composed of embedded-side and server-side.The embedded platform is only responsible for data acquisition and integrated navigation algorithms due to the limitation of computing power,while the server is responsible for data and model management.This dissertation opens with an explanation of the project background and research significance while investigates the related technologies and introduces the key theories.Then it makes a functional and non-functional requirements analysis for the system.Based on the requirements analysis,this dissertation designs an overall architecture for the system.And then the system is divided into several functional modules and each module is designed.At the same time,the detailed design of the system is completed,and the system is implemented by coding.Finally,the whole system is tested and analyzed based on the functional and non-functional.In this dissertation,the adaptive integrated navigation system is successfully implemented,and the system construction and test work are completed.The experiments based on practical road data prove the stability of the algorithm and the system,and the final navigation accuracy also fully meets the performance requirements.