基于MEMS传感技术的航姿系统算法研究

发布时间：2018-06-12 00:48

本文选题：MEMS + 航姿系统　；参考：《哈尔滨工程大学》2014年硕士论文

【摘要】：二十一世纪的高科技水平促进了微电子技术和微机械加工工艺的进步,在其推动下MEMS惯性传感器应运而生并且逐渐成为世界惯性领域中的研究热点。以MEMS惯性测量单元为主体的航姿参考系统以其微型化、轻质量、低成本等优势广泛应用于机器人、无人机、汽车、舰船、潜艇、通讯设备、生物医学等多个领域。微惯性系统由于其低精度和误差随时间累积等因素导致系统误差大且易发散,因此它不能够独立完成姿态测量工作。本研究在MEMS惯性测量单元的基础上以磁强计、GPS为辅助,融合三者信息形成优势互补的航姿系统。本文首先介绍了航姿系统中常用的参考坐标系以及磁强计测量原理,在此基础上研究了利用MEMS加速度计和磁强计输出信息进行解析式初始对准的方法。此外,介绍了 MEMS航姿系统的多种姿态更新算法并对各个方法进行对比分析,最终选择四元数法进行姿态更新。在基本理论学习的基础上本文对MEMS陀螺仪、MEMS加速度计以及磁强计进行误差分析并且分别建立了 MEMS陀螺仪输出误差模型、MEMS加速度计输出误差模型以及磁强计输出误差模型。然后,根据各误差模型分别设计了陀螺仪角增量标定方案、加速度计六位置标定方案,同时研究了基于椭球假设的磁强计误差标定方案,进而计算出各器件误差参数并进行相应补偿,为后续研究奠定坚实的基础。在航姿系统算法研究过程中首先建立了基于大失准角的MEMS惯性系统非线性误差方程,在有GPS信号的情况下,以针对非线性系统的EKF滤波算法进行MIMU、GPS、磁强计三者的信息融合,在此基础上利用GPS位置、速度和磁强计解算的磁航向与惯性系统自身解算的位置、速度、航向之差为EKF滤波的观测量,进而估计系统姿态,之后通过仿真验证本算法的有效性。在无GPS信号情况下,利用改进的基于四元数误差的自适应卡尔曼滤波算法实现MIMU、磁强计组合姿态估计。利用陀螺仪进行姿态解算的同时以加速度计和磁强计为辅助,提出了基于加速度计和磁强计的二阶量测更新算法,以及针对外部加速度的自适应算法,并且通过Matlab仿真验证了本算法的合理性。在仿真的基础上,分别设计了实验室静态试验、室外车载试验以及松花江江试试验。其中,静态试验结果较为理想;室外车载试验过程中汽车一直行驶在平稳路面,因此解算姿态具有一定的精度;而江试结果则进一步说明了在低动态条件下MIMU/GPS/磁强计组合方法能够保证姿态有效跟踪。
[Abstract]:The high technology level in 21 century has promoted the progress of microelectronics technology and micromachining technology. The MEMS inertial sensor has emerged as the times require and gradually become the research hotspot in the inertial field of the world. The attitude reference system based on MEMS inertial measurement unit is widely used in many fields, such as robot, UAV, automobile, ship, submarine, communication equipment, biomedicine and so on with its advantages of miniaturization, light weight and low cost. Due to its low precision and error accumulation with time, the micro inertial system can not complete the attitude measurement independently because of its large error and easy divergence. In this paper, based on MEMS inertial measurement unit, magnetometer GPS is used as the aid, and the three information are fused to form an attitude and navigation system with complementary advantages. In this paper, the reference coordinate system and the measuring principle of magnetometer are introduced. Based on this, the analytical initial alignment method using MEMS accelerometer and magnetometer output information is studied. In addition, various attitude updating algorithms of MEMS attitude and attitude system are introduced and compared with each other. Finally, the quaternion method is selected for attitude updating. Based on the basic theoretical study, the error analysis of MEMS accelerometer and magnetometer is carried out. The output error model of MEMS gyroscope and the output error model of MEMS accelerometer and magnetometer are established respectively. Then, according to each error model, the angular increment calibration scheme of gyroscope and the six-position calibration scheme of accelerometer are designed, and the calibration scheme of magnetometer error based on ellipsoid hypothesis is studied. The error parameters of each device are calculated and compensated accordingly, which lays a solid foundation for further research. The nonlinear error equation of MEMS inertial system based on large misalignment angle is first established in the research process of attitude and heading system. In the case of GPS signal, the information fusion of MIMUGPS and magnetometer is carried out for the EKF filtering algorithm of nonlinear system. On the basis of this, the magnetic heading calculated by GPS position, velocity and magnetometer and the position, velocity and course difference of the inertial system itself are used to estimate the attitude of the system by EKF filter, and the validity of the algorithm is verified by simulation. In the case of no GPS signal, an improved adaptive Kalman filter algorithm based on quaternion error is used to realize the combined attitude estimation of MIMUand magnetometer. By using gyroscopes to calculate the attitude of accelerometers and magnetometers, a second-order measurement updating algorithm based on accelerometers and magnetometers is proposed, and an adaptive algorithm for external acceleration is proposed. The rationality of the algorithm is verified by Matlab simulation. On the basis of simulation, laboratory static test, outdoor vehicle test and Songhua River test were designed. Among them, the static test result is more ideal, the vehicle has been driving on the smooth road during the outdoor vehicle test process, so the calculation attitude has certain precision; The result of Jiang test further shows that the combination of MIMUP / GPS / magnetometer can ensure effective attitude tracking under low dynamic conditions.
【学位授予单位】：哈尔滨工程大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN96;TP212

【参考文献】