Research On Cooperative Localization Of Multiple Mobile Robot In Satellite-jamming Environment

Since mobile robots have strong autonomy and adaptability to special working environments,they will assist or replace part of human work in the future,especially in battlefield mines and mine clearance,battlefield environment reconnaissance and scanning,disaster environment humanitarian search and rescue,terrain mapping,Geophysical prospecting data acquisition and other fields have extremely strong development potential.Multi-Mobile Robot System(MMRS)has advantages and potentials that the single mobile robot cannot match.Therefore,the related technology of MMRS has become an important development direction in robotics.In the working process of MMRS,the positioning accuracy of the system determines its operational performance.As the Global Navigation Satellite System(GNSS)has many advantages,such as high accuracy,wide coverage,and convenient operation,it is becoming an important means and preferred method for mobile robot positioning.However,because satellite signals are difficult to penetrate the ground and buildings,especially in indoor and underground environments,the GNSS fails and cannot be used.Therefore,how to realize the coordinated positioning of MMRS in the satellite rejection environment is the current focus for domestic and foreign research scholars.The inertial/visual integrated navigation system combines the advantages of these two systems and gradually becomes one of the main navigation methods of MMRS in the satellite rejection environment.Although the current MMRS technology based on inertial/visual integrated navigation has been developed rapidly,there are still many problems in complex environments,including low initial alignment accuracy caused by the complex uncertain systems,limited positioning accuracy resulted by low utilization of observation information,and low cooperative positioning accuracy leaded by less characteristic information in a non-cooperative environment,etc.This thesis has carried out an in-depth study on how to improve the cooperative positioning accuracy of MMRS in a satellite rejection environment.The main research contents are as follows:Firstly,in view of the problem that the nonlinearity and noise uncertainty of the MMRS in the satellite rejection environment affect the initial alignment accuracy,an improved high-precision initial alignment algorithm is proposed.The variance component estimation(VCE)method is used to adaptively estimate the noise characteristics of the system in real time,cubature Kalman filter(CKF)is used to avoid the truncation error caused by the linearization of the nonlinear system,and then the variance component estimation under extended Kalman filter(VCEKF)-based initial alignment algorithm is proposed.The inertial information and related constraints that are sensitive to the inertial sensor are used.A vector model based on the angular velocity of the earth’s rotation is established,and a latitude-free fast alignment algorithm based on gradient descent optimization is proposed.Simulation and experimental results show that the proposed improved initial alignment algorithm has good performance.Secondly,in order to solve the problem of insufficient use of measurement information during the cooperative operation of the MMRS,which limits the further improvement of the cooperative positioning accuracy,a cooperative localization algorithm based on a hybrid topology structure is proposed.The relationship between the observable matrix and the observation information is established,and an observability analysis method based on the relative position measurement graph(RPMG)is proposed.This method is used to conduct an observability analysis of the MMRS and the fully observabled conditions of the system are clarified.Under the extended Kalman filter(EKF)framework,a new cooperative localization algorithm based on the hybrid topology is proposed to improve the use efficiency of observations and cooperative positioning accuracy.The actual measurement data is used to verify the effect of the proposed cooperative localization algorithm.Thirdly,aiming at the problem that the cooperative localization algorithm of MMRS lacks initiative in a non-cooperative unknown environment,and lacks robust handling of observational uncertainty,which limits the accuracy of coordinated positioning,an adaptive active cooperative localization(A2CL)algorithm based on information optimization strategy is proposed.Carried out the optimization design of the control amount of the MMRS,and established the cooperative motion strategy based on model predictive control(MPC),which enables the robot to actively explore the environment information and maximize the amount of observation information.Further,in order to reduce the observation uncertainty effect of the cooperative positioning accuracy,the VCEKF method is used to estimate the covariance matrix of system noise and observation noise in real-time,improving the robustness and accuracy.The performance of the proposed algorithm is verified by semi-physical simulation experiments.The results show that the A2CL algorithm for MMRS based on MPC can not only effectively improve the cooperative positioning accuracy,but also has good robustness.Aiming at the problems that the initial alignment and the cooperative positioning accuracy of the MMRS cannot be guaranteed in the satellite rejection environment,the initial alignment algorithm and the cooperative localization algorithm for special application scenarios are respectively designed,further improved the cooperative positioning scheme for MMRS in the satellite rejection environment.