| With the development of science and technology,people's demand for location services is more urgent.The development of satellite navigation technology has become more and more mature,which has solved the problem of people's positioning in outdoor environment.However,in the complex and changeable indoor environment,satellite navigation is difficult to play its role.Among many indoor wireless positioning technologies,UWB signal uses sub-nanosecond wireless pulse to communicate.It has the advantages of high time resolution,anti-multipath,low power consumption and high ranging accuracy.UWB positioning accuracy can reach centimeter level in indoor LOS environment,but it is still affected by non-line-of-sight(NLOS)error,which leads to the decline of positioning accuracy.Inertial navigation system(INS)has the characteristics of autonomous positioning and is not affected by NLOS.It has high positioning accuracy in a short time,but the positioning error will accumulate gradually with the passage of time,so it needs to be corrected accumulative error.In order to restrain the influence of NLOS error in UWB positioning,the indoor positioning method based on UWB/INS data fusion can make use of the positioning advantages of different subsystems to complement each other,and can meet the requirements of reliability and continuity of indoor positioning.The main research work is as follows:(1)Aiming at the measurement error problem of the combined positioning system,the data preprocessing is carried out in this paper.Linear fitting of UWB ranging value corrects system error and achieves centimeter-level ranging accuracy;the constant drift error of the gyroscope is corrected so that the corrected error is close to zero;the accelerometer and magnetometer were calibrated by ellipsoidal fitting.The standard deviation of gravity field and geomagnetic field modulus measured before and after calibration was reduced from(0.0111 g,0.1053Gauss)to(0.0022g,0.01862 Gauss),respectively.the measurement accuracy of INS was improved.(2)In order to solve the problem of attitude detection error and rotation matrix error caused by magnetic interference and external force interference,the weighted suppression is carried out in this paper.In the quaternion attitude detection algorithm,the original two-step updating algorithm only avoids the influence of magnetic interference on horizontal tilt angle(roll angle,pitch angle),but the influence of magnetic interference on heading angle and external force interference on horizontal tilt angle has not been suppressed.On this basis,the weighted piecewise function is used to restrain the influence of external disturbance on attitude angle.Experiments show that the algorithm has a certain suppression effect on interference.(3)In combination algorithm,aiming at the problem that centralized EKF processing data makes the system dimension larger and Jacobian matrix solving process complex,this paper designs a cascade EKF data fusion form.The first EKF is used for attitude detection and rotation matrix solution,the second EKF is used for position calculation,and the velocity information in the system state is updated and corrected by the historical trajectory of combined positioning.The combined system adopts the tight coupling method to determine the NLOS error and the anomalous value of ranging based on the absolute value of ranging residuals,and uses the UWB ranging value under the line-of-sight to correct the cumulative error of INS prediction position,so as to ensure the continuity and reliability of the positioning system.The experimental results show that UWB/INS combined positioning algorithm can effectively suppress indoor NLOS errors.When NLOS errors occur,the 95%probability positioning error of the combined system is less than 0.07m under static conditions.Under motion conditions,compared with the reference trajectory,the maximum trajectory deviation is only 0.35m,and the overall positioning error is basically less than 0.2m. |
PDF Full Text Request