Linearization Initial Alignment Under Large Misalignment Angles For SINS

Initial alignment technology is a key research point for strapdown inertial navigation system(SINS),which directly affects the main performance of SINS.By taking the various grades of SINS as the research object,self-alignment under large misalignment angles,semi-self-alignment based on azimuth setting,the transfer alignment in inertial frame and GPS-aided at random misalignment angles are discussed.The contents involve the state-matrix Kalman filter,the attitude determination algorithm,the SINS error model and the filtering algorithm.The main contents of the research include:(1)A state-space model and Kalman filter in matrix form are derived and summarized.The general form of state-matrix Kalman filter is derived which operates on plants described naturally by matrix processes and matrix measurement equations.It shows that the matrix Kalman filter(MKF)is a direct generalization of the classical discrete vector and Kalman filter Kalman filter.Operating on this plant the MKF,the raditional nonlinear model under large misalignment angles is modelled as a linear state matrix alignment model.In this work,this paper designs a linear alignment algorithm without initial parameters and distinguishing between coarse and fine alignment processes in inertial frame.(2)Self-alignment algorithm under large misalignment angles is researched.Based on the principle of inertial frame alignment and taking the constant matrix as the estimation object,two kinds of nonsingular linear alignment model under large misalignment angles are given: the K-matrix Kalman filter alignment algorithm,and SO(3)constraint-based Kalman filter alignment algorithm.With the K-matrix and vectorized attitude matrix as the estimated states,the system equations and measurement equations are linear equations,yields,the traditional nonlinear initial alignment model under a large misalignment angle is converted into two kinds of linearization alignment model in inertial frame.A new latitude estimation algorithm and initial alignment scheme with unknown latitude are presented,and the algorithm is verified by vehicle alignment test.For the low-cost INS,the semi-self-alignment scheme of attitude self-alignment and azimuth setting is designed.the alignment result is attitude quaternion,simultaneously,the alignment process can also estimate gyro drift roughly.(3)A transfer alignment(TA)scheme based on matrix Kalman filter framework in inertial system is designed.Analyzing the measurement vector matching,and employing the relative attitude matrix between the master INS(MINS)and slave INS(SINS)as estimate state,then a unified nonsingular linear transfer alignment model is designed,and the specific force integration plus angular rate integration in inertial system is constructed as the matching quantities.Providing the detailed implementation of Matrix Kalman Filter(MKF),and verifyinf the effectiveness of the algorithm by simulation under the sway maneuver.Finally,a TA accuracy evaluation algorithm based on the RTS is presented.(4)A in-motion alignment algorithm of SINS/GPS based on filtering scheme at arbitrary misalignment angles in inertial frame is investigated.Firstly,the principle of the initial alignment algorithm based on GPS and the formula of measurement calculation are deduced in detail.With the idea of inertial frame alignment,the alignment problem is converted to the estimation of the attitude matrix at the initial alignment time.On this basis,the rotation vector,quaternion multiplicative extended Kalman filter alignment algorithm and attitude matrix SO(3)-constrained Kalman filter alignment algorithm are designed.Finally,the vehicle experiment is designed using the MEMS IMU/GPS system to validate the algorithm.(5)The application of SINS in precise airdrop is investigated.The particularity of precise airdrop system in-flight alignment and the error sources of GPS-aided alignment are analyzed.On the basis,attitude matrix the SO(3)-constrained Kalman filter alignment scheme is proposed.The navigation error is better than 200 meters within 120 seconds after IMU error compensation.Trajectory simulations,vehicle experiment and actual airdrop test are performed to validate that the algorithm is verified.Finally,the path planning of the airdrop system is studied.The determination of the airdrop area and the real-time wind estimation are described in detail,and the turning track based on Durbins curve is planned under the constraints of the shortest flight path.