Research On Initial Alignment Of Land-Vehicle Strapdown Inertial Navigation System In-Motion Aided By Odometer

The purpose of the initial alignment is to establish the initial navigation coordinate for the strapdown inertial navigation system(SINS).The navigation accuracy is determined by the accuracy of the initial alignment,and the response time of the vehicle-based weapon is determined by the speed of the initial alignment.In order to effectively enhance the maneuverability of the vehicle-based weapon,improve the combat capability and survival ability,it is necessary to achieve high precision alignment in-motion.In this paper,the research focuses on the odometer aided initial alignment of land-vehicle SINS in-motion,and the system error model,observability analysis,initial alignment algorithm based on optimal estimation and nonlinear filtering are studied.Aiming at the initial alignment of large misalignment angle,the initial alignment model of SINS/Odometer system is established in detail and the odometer/gyroscope dead reckoning(DR)scheme is adopted.Firstly,the SINS error equations of small and large misalignment angle are derived.Then,based on the SINS attitude error equation,the odometer/gyroscope dead reckoning error equations are deduced and established.With the DR scheme,a liner initial alignment measurement equation is established,which can simplify the measurement equation of the SINS initial alignment and has not been fully explored before.In addition,a high-precision lever arm model and dead reckoning algorithm are proposed to improve the navigation and alignment accuracy on coarse road.The observability of the SINS/Odometer system is analyzed before designing the initial alignment scheme.In order to avoid the disadvantages of the PWCS(Piece-Wise Constant System)observability analysis method,which takes a huge amount of calculation and has linearization error for high dimensional and highly non-linear system.This paper employs the global observability analysis method,starting from the concept of system observability,exploring the system dynamic equation,and turning the problem that whether the system state including attitude,speed,IMU bias,SINS installation misalignment angle,odometer scale factor and the lever arm,can be observable or not into a problem that whether they can be solved by measurement during limited time or not.This method not only reaches the conclusion that the system is observable,but also reaps the benefits of the sufficient condition of the system state to be observable,providing the theoretical guidance for the initial alignment maneuver scheme.Based on the idea of “inertial frame based alignment”,an optimization-based initial alignment and calibration algorithm of land-vehicle SINS in-motion is developed.Firstly,an on-line integration calibration scheme for the SINS installation misalignment angle and odometer scale factor is designed according to the results of observability analysis.Then,the initial alignment problem is transformed into a constant attitude determination problem through the decomposition of attitude matrix,the initial alignment numerical integration algorithm is derived,and the initial alignment scheme based on constraint optimization is proposed.The scheme can online estimate system states including attitude,IMU bias,SINS installation misalignment angle,odometer scale factor and lever arm.Different from the Kalman filter method,this method does not require the knowledge of the statistical characteristics of system noise,and is more robust and has stronger anti-interference ability than the EKF method.The vehicle initial alignment test indicates that by using the proposed OBA method,the in-motion initial alignment accuracy can be dramatically improved,which demonstrates the superiority of the proposed method.To solve the problem of divergence and accuracy reduction of standard Kalman filter caused by non-linear model of low precision SINS/Odometer system and uncertainty of measurement noise,an improved ACKF/KF initial alignment method is presented.In this method,the nonlinear filtering algorithm – cubature Kalman filter is used to handle the nonlinear system,and the Sage-Husa adaptive filter algorithm is adopted to estimate system measurement noise online.And according to the linear system measurement equation obtained by the odometer/gyroscope dead reckoning,the standard Kalman filter is used to carry out measurement update,which reduces the calculation amount of filter and relieves the "dimensional disaster" of the nonlinear filter.Therefore,the proposed ACKF/KF algorithm can perform state estimation for the nonlinear SINS/Odometer initial alignment equations with unknown measurement noise,and has better engineering practicability.The ACKF/KF algorithm is compared with EKF,CKC and AEKF through initial alignment simulations,the results show that the ACKF/KF algorithm can effectively improve the alignment accuracy and stability,and can better complete the SINS initial alignment in-motion.