Research Of Calibration And Initial Alignment For Strapdown Inertial Navigation System

Strapdown inertial navigation system(SINS)is the research emphasis in the field of inertial navigation.The calibration and initial alignment of SINS are the critical processes performed prior to normal navigation.To improve the related technologies of them,several researches have been conducted in this paper,including the least square fit(LSF)method of systematic calibration,coarse alignment based on inertial reference frame,and gyrocompass alignment using double systems.The main research works of this paper contain:1.In order to reduce the influence of the measurement noise,the LSF method of navigation-in-velocity is modified based on the stationary condition and an improved LSF method using specific force and angular velocity as observations is presented.Then,observability analysis of the improved method is made.After that,we study the influence of rotary mechanism errors on this method.The theoretical analysis and simulation results show that the impact of the errors could be eliminated by iterative method as the magnitudes of them within a certain range.2.Without the limitation of inertial measurement unit(IMU)orientation,a systematic calibration method based on the stationary base is provided,and it reduces the requirements of the horizon and north.Then with the help of the linear algebra,an observability analysis of the proposed method is made,and the requirement of completely observation is given under the classical rotary indexing sequence.On the basis of the observability analysis,we analyze the effect of gravity anomaly on the proposed method,and confirm the calibration parameters which are influenced by the gravity anomaly.After that,the impact of IMU measurement noise on this method is analyzed,and the relationships between measurement noise and calibration parameters are provided.In order to simplify calculation,iterative computational method is derived.Simulation and experiment results validate the above methods and analysis conclusions.3.Combined with the analysis of the disturbance intensity under the base motion,we analyze the existent problems of the calibration method proposed in Chapter 3.After that,in order to solve the application issue of motion base,we design a systematic calibration method by means of kalman filter(KF)and LSF.On the basis of the above contents,we study the effect of measurement noise and motion disturbance on the proposed method.And then,a unified algorithm is designed under the classical rotary indexing sequence,to improve the applicability of the proposed calibration method.Simulation and experiment results validate the above methods and analysis conclusions.4.The coarse alignment based on inertial reference frame is studied,and we mainly take velocity-type and position-type coarse alignments as research objects.Firstly,the construction principle and error sources of these two methods are analyzed,and then simulation results demonstrate the conclusion which the influence of sensor errors on the two methods is certainty.Consider the motion disturbance,we derive the design principle of optimal parameters.Based on the principle,the optimal parameters of the coarse alignments are obtained.Finally,the performances of velocity-type and position-type coarse alignments are compared with each other.The analysis conclusions are verified by simulations and experiments.5.With the development of navigation computer,several strapdown computational algorithms can be utilized during its operation simultaneously.According to this characteristic,a gyrocompass alignment method using double systems is presented.On the basis of this method,an improved method is designed to solve the nonlinear problem of alignment model under the large azimuth misalignment,and the alignment duration is reduced.To get a high accurate steady alignment result,we study the error characteristics of the improved method under the rotary modulation,and point out the approach to improve the alignment performance.Simulation and experiment results validate the above methods and analysis conclusions.