摇摆基座下旋转捷联系统粗对准技术研究

阐述了利用惯性测量单元(inertial measurement unit, IMU)的转动调制惯性器件常值偏差的原理.针对载体处于摇摆基座下难以实现粗对准这一问题,在已有惯性系对准方法的基础上提出了改进的惯性系对准方案并应用于旋转捷联惯导系统中.利用数字低通滤波器滤除由于摇摆和振荡运动产生的加速度干扰,实现了旋转捷联系统的粗对准.对比地分析了2种惯性系下的粗对准原理并进行了仿真,利用系泊实验进一步验证了改进的惯性系对准方案的可行性.结果表明,载体处于摇摆状态时,采用低通滤波方法可以有效地提取基座惯性系下的重力加速度信息,并建立更为准确的捷联矩阵;与传统的惯性系对准方法相比较可以看出,采用滤波技术的惯性系对准结果具有更好的稳定性,存在广阔的应用前景.     

加权递推最小二乘的捷联式惯导系统初始对准方法研究

捷联式惯导系统(捷联式惯性导航系统)的初始对准就是确定捷联矩阵的初始值,陀螺与加速度计的误差会导致对准的误差,对准时飞行器的干扰运动也是产生对准误差的重要因素,因此滤波技术对捷联式惯导系统尤为重要.通过研究捷联惯导初始对准的误差模型,提出了将加权递推最小二乘滤波算法应用于捷联式惯导系统的初始对准中,仿真结果表明,该方法估计精度高,收敛速度快.     

一种单轴旋转捷联惯导的三位置对准方法

针对单轴旋转式捷联惯性导航系统,提出了一种基于惯性测量单元(inertial measurement unit,IMU)旋转的三位置初始对准方法.在系统可观测性分析的基础上,建立IMU姿态与失准角之间的关系,提出了使失准角估计偏差为零的三位置对准方法.与以往固定位置对准和两位置对准进行了比较,表明该方法可以消除不可观测的水平加速度计零偏和东向陀螺常值漂移对初始对准精度的影响,大大提高了系统初始对准精度.     

卡尔曼滤波技术在捷联惯导系统初始对准中的应用

阐述了卡尔曼滤波技术在捷联惯导系统初始对准中的应用,建立了捷联惯系统初始对准的误差模型和卡尔曼滤波模型观测方程,进行了卡尔曼滤波仿真,并分析了提高采样频率对系统的估计精度和收敛速度的影响。仿真结果表明,此方法算法简单,能有效缩短初始对准时间,对准稳态精度较高,是一种理想的初始对准方案。     

捷联惯导系统极区水下动基座对准EI北大核心CSCD

惯性导航系统正常工作前需进行初始对准,极区和水下环境中的动态对准是惯导系统初始对准的难点。通过惯性坐标系下的惯导系统误差模型进行卡尔曼滤波初始对准,避免了对准误差随纬度升高而发散的问题。在水下利用载体坐标系测速设备提供速度参考,为卡尔曼滤波器提供速度误差观测量。在对准过程中,利用航位推算方法对载体位置进行实时更新,进一步提高了对准精度。通过跑车数据及仿真试验对算法进行验证,结果表明,所提算法可以有效实现惯导系统在极区水下环境中的动态初始对准。     

适用于低精度惯导的非线性对准方法研究

给出了一种适用于低精度惯导的非线性对准模型.用乘性四元数形式定义捷联惯导的姿态误差,推导了捷联惯导的非线性速度误差方程和姿态误差方程.基于速度量测信息,给出了大失准角条件下的非线性对准模型,通过UKF算法估计失准角完成精对准.仿真结果表明,在陀螺精度为0.1°/h的情况下,在360s对准时间内达到水平0.03°,方位1.5°的精度(1σ).即使当方位误差达到90°,非线性模型仍能正常收敛.最后通过转台摇摆试验进一步验证了非线性模型的有效性.     

适用于强阵风干扰下的捷联惯导自对准算法

对于处于起竖状态的车载武器来说,在强阵风的干扰下会产生较大摇晃,由于其捷联惯导系统(SINS)的重心较高,SINS的测量值很容易受到晃动所带来的角运动和线运动干扰,SINS难以快速地实现自对准,针对该问题,提出了强阵风干扰下SINS自对准算法。该算法利用惯性坐标系下的姿态更新来实时地反映载体在晃动干扰下的姿态变化,以消除角运动干扰的影响;利用武器存在零速摇晃中心的特点,通过获取加速度计在零速摇晃中心的等效输出来消除线运动干扰的影响,然后结合姿态的最优估计实现初始对准。仿真结果表明,该算法不需要进行粗对准,能够在强阵风干扰下快速地实现自对准。     

基于CKF的SINS大方位失准角初始对准

大方位失准角捷联惯导系统(strapdown inertial narigation system,SINS)误差方程是非线性的,为改善非线性模型下初始对准的精度,提出将一种新的非线性滤波方法(cubature Kalman filter,CKF)应用于捷联惯导系统初始对准中。为此建立了大方位失准角下初始对准非线性模型,分析了基于spherical-radial cubature准则的CKF滤波原理,对非线性模型进行了CKF滤波仿真。仿真结果表明CKF能够很好地处理初始对准中的非线性问题,提高初始对准精度,方位失准角误差能够收敛到-33.13’,接近理论估计精度-32.40’,优于EKF的-84.14’。     

QUKF算法及其在SINS初始对准中的应用

针对捷联惯导系统初始对准过程中的大失准角情况,建立了乘性四元数姿态误差方程。结合UKF算法,提出了基于四元数的UKF算法(QUKF)。以姿态矩阵为对象,通过构造姿态矩阵代价函数的方法来求取四元数的一步预测均值,保证了均值四元数的规范化;利用乘性误差四元数表示四元数Sigma点与均值四元数的距离,以求取四元数的预测协方差矩阵。四元数状态更新中,采用乘性四元数更新保证了滤波过程中四元数的规范化。基于该算法的SINS在粗对准水平失准角为小角度、方位失准角为大角度条件下的仿真实验结果表明,与常规EKF相比,纵、横摇角对准精度均略有提高,而航向角误差估计精度提高显著。     

惯导系统误差仿真

利用静基座下惯性导航系统的误差方程，应用Matlab／Simulink建立了惯导误差仿真模型。文中对初始值误差，加速度计零位误差及陀螺常值漂移共同作用及各自分别作用条件下进行了仿真，通过仿真曲线对各误差源进行了简要地分析。     

晃动基座下激光陀螺捷联导航系统初始对准预滤波新方法

激光陀螺捷联惯性导航系统在晃动基座下进行初始对准,外界的干扰会令激光陀螺和加速度计的噪声增大,从而使初始对准的时间延长,甚至不能完成对准。对激光陀螺和加速度计输出简单的采用低通滤波的方法,并不能有效的抑制传感器噪声。采用小波方法,虽然能够抑制噪声的影响,但是由于小波的高计算复杂度和块处理结构,很难在线实现。针对这个问题,本文提出一种新的预滤波方法,将传感器输出通过低通滤波之后再通过一个基于隐式马尔可夫模型的稳态卡尔曼滤波,这样就能有效的降低基座晃动带来的噪声,同时不降低对准的精度。实验结果表明,本文提出的预滤波方法计算复杂度低,滤波效果明显,能够很好的辅助激光陀螺捷联导航系统在晃动基座下完成精对准。     

Initial alignment on moving base using GPS measurements to construct new vectors

Strapdown inertial navigation system cannot easily implement initial alignment on moving base if without the prior “one shot” transfer from another master inertial navigation system and strict constraints on moving trajectory, because the initial coarse attitude cannot be effectively achieved. In this paper, a novel method using the GPS measurements to assist to solve initial attitude matrix is investigated as coarse alignment. Specific vectors are constructed with GPS measurements and used for attitude matrix computation like in analytic alignment algorithm. Several moving conditions of carrier are assumed to test the new alignment method, and the results indicate that the initial attitude angles can be achieved with small errors, which is suitable for the conventional fine alignment afterward.     

MEMS-based rotary strapdown inertial navigation system

As the inertial navigation completely depends on the sensed acceleration and rotation rate by IMU, the sensor errors accumulate and eventually lead to diverged inertial solutions. Therefore the compensation of the inertial sensor errors is an effective approach to improve the SINS navigation performance. The rotation error modulation in rotary SINS, which has been extensively used for the filter-optical IMU in the past, is one of the techniques to compensate or mitigate the inertial sensor errors and eventually improve the system navigation performance. The rotary SINS is an inertial navigator in which the IMU is installed on the rotational platform and rotated following the pre-designed rotation configuration, and the rotation error modulation is the technique that compensates the navigation errors caused by inertial sensor bias in a complete rotation cycle by rotating IMU. Given the auto-compensation of inertial sensor bias in rotation error modulation, the objective of this paper to develop a MEMS-based rotary SINS, in which the significant sensor bias is automatically compensated by rotating the IMU, to offer the comparable navigation performance to tactical-grade IMU. Simulation results indicate that, compared with the conventional method, the proposed approach attenuates the navigation errors and improve the calibration accuracy based on the reciprocating rotation scheme can be used to automatically improve the observability.     

机载光电/惯性组合着舰导引算法研究

提出了一种基于红外探测的机载光电/惯性组合着舰导引新方案,综合机载红外热像仪和惯导信息实现舰、机运动分离,完成舰载机半自动或自动着舰导引,具有精度高、自主性好、不易受电磁干扰的优点.阐述了系统方案,并深人研究了靶标规则摆放情况下的着舰导引算法,给出了计算机仿真结果.仿真结果表明,舰机相对航向角估计误差小于10′、着舰位置估计误差小于1 m,验证了方案的可行性及算法的有效性,具有较高的工程应用价值.     

混合式惯导系统自标定技术研究

混合式惯导系统是一种集平台式、捷联式、旋转式惯导系统优点于一体的新型惯导系统。该型惯导可利用自身的旋转机构和高精度、高分辨率角度传感器实现不拆机条件下的误差参数自标定功能,极大简化了标校流程和系统维护工作量,有利于武器平台和作战系统效能的发挥。针对混合式惯导系统的结构特点和典型的旋转方式进行了误差分析,给出了误差参数自标定的设计原则和方法,并用混合式惯导原理样机进行了验证,结果表明,所设计的自标定方案能够在不拆机条件下利用自身旋转机构完成对陀螺漂移、加计零偏、刻度系数误差及安装偏角等参数进行精确估计,因此具有较高的理论意义和工程应用价值。     

长航时惯导系统全阻尼综合校正算法

针对惯导系统长时间工作时导航误差随时间发散的问题,提出一种适用于长航时惯导系统的全阻尼综合校正算法。设计了全阻尼网络,利用外部速度量测对惯导系统进行水平和方位阻尼,衰减了导航误差中舒拉周期及地球自转周期振荡;根据全阻尼条件下的惯导系统误差传播规律设计了综合校正算法,利用不定期获取的外部参考位置计算导致导航误差发散的陀螺漂移并进行修正,更好地保证了惯导系统的工作自主性。试验结果表明,该算法明显地抑制了惯导误差随时间的发散,可有效提高长航时惯导系统的导航精度。     

An initial alignment method for strapdown gyrocompass based on gravitational apparent motion in inertial frame

Inspired by the alignment mechanism of Octans, an initial alignment method for strapdown gyrocompass is designed based on gravitational apparent motion in inertial frame. In this method, the gyro outputs are used to trace the body frame and the measurement from accelerometers are projected to the inertial frame to form the track of gravitational apparent motion. According to the calculations on the vectors involved in this track, the attitude matrix between the inertial frame and navigation frame can be obtained and further alignment can be finished. In order to identify the true gravitational apparent motion from the accelerometer measurement containing random noise as well as to avoid the collinear of vectors in vector operation, a reconstruction algorithm concerning parameter identification and reconstruction of apparent motion is devised and simulated. Simulation and turntable test show that the proposed alignment method can realize self-alignment in a swinging condition. The alignment accuracy is able to reach the theoretical one determined by sensor errors and no external information is needed.     

A fast compass alignment method for SINS based on saved data and repeated navigation solution

Compared with Gimbaled Inertial Navigation System (GINS), there are two differences in the compass alignment process for Strapdown INS (SINS) as follows: (1) the input signal is the projected sensor data from body frame to calculated navigation frame, not the sensor data; (2) the output control signal participates in navigation solution rather than driving navigation frame directly. Considering these characteristics, a fast compass alignment method based on iterative calculation is designed for SINS in this paper. With the help of high performance computer, the iterative process can be finished fast and the alignment time can be shortened obviously. However, there exist two problems in the new method, one of which is how to use the saved data, and the other is how to ensure the real-time property of alignment result. For the first problem, our simulation result shows that the forward–forward navigation solution method is more effective for compass alignment than the backward–forward one. And to solve the second problem, Multi-tasking method combined with Real Time Multi-tasking Operating System (RTOS) is used to trace the change of body frame and update the body frame of attitude matrix when the alignment is finished. The car test on a type of fiber-optic gyro SINS indicates that the alignment accuracy of the proposed method processed for 142 s is as precise as that of the classical compass alignment method for 720 s.     

无人机磁惯导系统中陀螺仪的叉积标定算法*

针对无人机磁惯导系统(MINS)中的三轴MEMS陀螺仪的现场标定问题,提出了一种叉积标定算法。该方法基于载体矢量变化率与角速度关系来标定陀螺仪的各项误差,即导航坐标系中任一常矢量在无人机载体坐标系中对时间的变化率,可表示为该矢量与载体角速度的叉积。叉积标定法利用这一原理,可在无精密设备的条件下快速、便捷地对三轴陀螺仪进行标定。数值仿真表明,叉积标定法在微分形式和积分形式下都能有效地辨识和补偿陀螺仪的各项误差系数,并且能在各种不同因素影响下获得良好的标定效果。实物试验结果证明,基于地磁矢量辅助的叉积标定法对MINS模块中陀螺仪进行标定时,其精度达到0.227 9°/s,能够接近于转台标定的水平。陀螺仪标定后的数据与旋翼无人机飞控中的二阶互补滤波算法相结合,使得定点悬停状态下角度偏差控制在0.8°以内,有利于无人机现场标定和实际飞行中姿态数据的量测。