基于多传感器信息融合的低成本车载组合导航的研究

发布时间：2018-05-07 03:09

本文选题：MIMU/WSS/MAG组合导航系统 + 初始对准　；参考：《山东大学》2017年硕士论文

【摘要】：当使用低成本微惯性测量单元(MEMS IMU)作为惯导系统的惯性测量元件时,受到惯性传感器精度限制(主要表现为陀螺和加速度计的漂移和噪声),纯惯导系统无法独立完成导航定位,将其与其他导航方式结合成为了陆用车载导航的主要研究方向。同时,传感器信息融合技术的发展也使得高度组合的车载自主导航系统的实现成为可能。本文以车载组合导航系统为应用背景,重点研究了低成本MIMU/WSS/MAG组合系统的关键技术,主要研究内容包括:1.推导了捷联惯导系统(Strapdown Inertial Navigation System,SINS)姿态、速度和位置更新算法,其中姿态更新采用四元数更新并在载体坐标系更新过程中考虑到圆锥运动误差,在速度更新过程中考虑划船误差补偿。推导了捷联惯导系统误差状态简化模型,以便用于低成本组合导航系统设计。最后通过MIMU在静态情况下的测量值,进行纯惯导解算,导航结果表明当使用MIMU作为惯性测量元件时,纯惯导系统的在前1Os内有很好的精度,但之后的姿态、速度和位置误差会快速增大,说明设计组合导航系统的必要性。2.为了解决组合导航系统的初始对准问题,本文研究了磁强计(MAG)辅助的初始对准算法。首先针对复杂的车载环境,通过分析磁强计所受到的软磁和硬磁干扰,采用椭球拟合来补偿磁强计受到的罗差干扰。通过对比补偿前后磁强计输出的磁场模值,补偿测得的磁场模值接近1,证明很好地补偿了磁强计的罗差。然后利用加速度计和磁强计实现了粗对准,为精对准提供了较为精确的初始姿态角,而且大大地缩短了粗对准的时间。最后,设计了静基座下基于卡尔曼滤波理论的初始精对准算法。静基座下的初始对准实验表明本文提出的初始对准算法能够在很短的时间内收敛,对准精度满足惯导的要求。3.设计并且实现了 MIMU/WSS/MAG车载组合导航系统。由于联邦卡尔曼滤波器相对于集中式卡尔曼滤波具有计算量少、便于对导航子系统作故障检测、诊断和隔离等优点,本文详细描述了联邦卡尔曼滤波器的设计过程。在仔细分析子导航系统组合模式后,分别设计了 SINS/WSS和SINS/MAG子系统,建立各子系统的状态方程和观测方程,然后根据联邦卡尔曼滤波的自适应信息分配原则将两个子系统进行了融合,完成了自适应联邦卡尔曼滤波器的设计。最后设计了实车实验,通过对比联邦卡尔曼滤波和集中式卡尔曼滤波的导航结果,验证了联邦卡尔曼滤波在导航精度和故障隔离及恢复的优势。论文对低成本MIMU/WSS/MAG车载组合导航系统进行了较为系统的研究,为低成本车载组合导航提供了一种可行的组合方案,为进一步实现低成本高度融合的车载组合导航系统奠定了基础。
[Abstract]:When MEMS IMU is used as the inertial measurement element of inertial navigation system, it is limited by the precision of inertial sensor (mainly the drift and noise of gyroscope and accelerometer, so the pure inertial navigation system can not complete navigation and positioning independently. Combining it with other navigation methods has become the main research direction of vehicle-borne navigation. At the same time, the development of sensor information fusion technology also makes it possible to realize the highly integrated vehicle autonomous navigation system. In this paper, the key technology of low cost MIMU/WSS/MAG integrated system is studied in the background of vehicle integrated navigation system. The main research contents include: 1. An algorithm for updating attitude, velocity and position of Strapdown Inertial Navigation system sins is derived, in which quaternion updating is used in attitude updating and conical motion error is taken into account in the updating process of carrier coordinate system. In the process of speed updating, the compensation of rowing error is considered. The simplified error state model of strapdown inertial navigation system is derived for low cost integrated navigation system design. Finally, through the measurement value of MIMU in static condition, the pure inertial navigation solution is carried out. The navigation results show that when MIMU is used as the inertial measurement element, the pure inertial navigation system has a good precision in the pre-1Os, but the attitude of the pure inertial navigation system after that. Speed and position errors will increase rapidly, indicating the need to design integrated navigation system. 2. In order to solve the problem of initial alignment of integrated navigation system, the magnetometer aided initial alignment algorithm is studied in this paper. Firstly, aiming at the complex vehicle environment, by analyzing the soft magnetic and hard magnetic interference of the magnetometer, the ellipsoid fitting is used to compensate the difference interference of the magnetometer. By comparing the magnetic field mode values of the magnetometer before and after the compensation, the measured magnetic field mode value is close to 1, which proves that the error of the magnetometer is well compensated. Then the coarse alignment is realized by using accelerometers and magnetometers, which provides a more accurate initial attitude angle for fine alignment, and greatly shortens the time of coarse alignment. Finally, the initial fine alignment algorithm based on Kalman filter theory is designed under the static base. The initial alignment experiments under the static base show that the proposed initial alignment algorithm can converge in a very short time and the alignment accuracy can meet the requirements of inertial navigation. MIMU/WSS/MAG vehicle integrated navigation system is designed and implemented. Because the federated Kalman filter has the advantages of less computation than centralized Kalman filter, it is convenient to detect, diagnose and isolate the navigation subsystem, the design process of the federated Kalman filter is described in detail in this paper. After carefully analyzing the integrated mode of subnavigation system, the SINS/WSS and SINS/MAG subsystems are designed, and the state equations and observation equations of each subsystem are established. Then the two subsystems are fused according to the adaptive information allocation principle of federated Kalman filter and the design of adaptive federated Kalman filter is completed. Finally, a real vehicle experiment is designed, and by comparing the navigation results of federal Kalman filter and centralized Kalman filter, the advantages of federated Kalman filter in navigation accuracy, fault isolation and recovery are verified. In this paper, the low cost MIMU/WSS/MAG vehicle integrated navigation system is systematically studied, which provides a feasible integrated scheme for low cost vehicle integrated navigation system, and lays a foundation for the further realization of low cost and high integration vehicle integrated navigation system.
【学位授予单位】：山东大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U463.675

【参考文献】