The Research About Airborne Vector Gravimeter And Methods Of Errors Separation

Airborne gravimetry is an effective method to collect the homogeneous distribution, precise and large area information about the Earth’s gravity. Strapdown airborne gravimeter has many advantages of being simple structure, small size, light weight, low cost, low power consumption. Therefore, airborne gravimeter based on SINS has been the trend of development of the airborne gravimetry. Beside this, strapdown gravimeter is also able to implement vector gravimetry, and the gravity vector information is significant in many subjects such as geophysics.The researches about this paper are based on a strapdown airborne gravimeter SGA-WZ, which was developed by the National University of Defense Technology. The thesis carried out in-depth researches in strapdown airborne vector gravimetry error analysis, the influence on vector gravimetry from inertial device error and the error compensation, the influence on vector gravimetry from gravity disturbance, the influence on vector gravimetry from the dynamics of aircraft and the error separation from the gravity vector results. Main results are summarized as follows:1. It carried out in-depth study in the principle of strapdown airborne vector gravimetry. Thesis discussed the method of gravity vector measurement based on the position update, including mathematical model and error model. According to the error model of airborne vector gravimetry, it designed an orthogonal experiment to analyse the influence degree of each factor for airborne vector gravimetry.2. Aiming at the inertial device error estimation and compensation, the IPSE algorithm was proposed. The 15-states Kalman filter was often used to improve the accuracy of specific force in the data processing. However, constant speed and straight course makes ordinary 15-states Kalman filter does not meet the observability conditions during the airborne surveys. Thesis analyse the observability of the 15-states filter and point out that the accelerometer bias and the attitude error can not be estimated respectively, only the linear combination of them can be estimated. The IPSE algorithm proposed in the paper made the process of device error estimation much more clear. Simulation and experiment show that the IPSE algorithm can effectively obtain a good estimation of the device error. Experimental results show that the accuracy of north component of gravity vector improved from 15.57 m Gal to 4.12 m Gal and the accuracy of east component of gravity vector improved from 17.36 m Gal to 14.70 m Gal.3. The existence of the gravity disturbance will affect the gravity vector measurement. Simulation shows what influence will be induced by using normal gravity in the navigation compution. In order to solve this problem, the iteration algorithm was researched. The simulation results testify the validity of the iteration algorithm and point out the working conditions of the iteration algorithm. Experimental results shows that the iteration algorithm can make the diffence between the gravity estimation and the real gravity imformation converge to a constant value. After the direct current components were wiped off, the accuracy of north component of gravity vector improved from 4.12 m Gal to 2.57 m Gal and the accuracy of east component of gravity vector improved from 14.70 m Gal to 3.06 m Gal.4. The method of airborne gravimetry is determining gravity using instruments taken on an airborne vehicle, therefore, the dynamics of the aircraft will affect the gravity vector measurement. Thesis points out that the phugoid modes and the slow spiral modes will influence the airborne gravimetry possibly, and the slow spiral modes only induce the positing error. Therefore, the phugoid modes will bring material influence to the airborne gravimetry. Two practical airborne gravimetry campaigns were used in order to illuminate the real dynamics of aircraft, and the evaluation index was also proposed. According to the error spectrum of the gravity vector measurement, thesis discussed how to reduce the measurement error induced by dynamics of aircraft and implement the synchronization error and attitude error estimation and compensation. At last, the influence of the dynamics of aircraft was suppressed using EMD.5. There are errors in the gravity vector results. In order to solve this problem, thesis researched the error separation method. Firstly, the Wavenumber Correlation Filter was proposed, this method is suitable for repeat lines or parallel lines nearby. The basic hypothesis of the WCF is that the gravity disturbance from different survey lines that are on the same trajectories should be highly correlated, while the systematic errors should be less correlated in the opposite directions. The error will be separated by suppressing the low correlated components. Experimental results show that the accuracy of north component of gravity vector improved from 2.49 m Gal to 1.43 m Gal and the accuracy of east component of gravity vector improved from 2.81 m Gal to 1.80 m Gal after WCF. Secondly, the application of the backward navigation computation of Kalman filter was researched. The errors were wiped off by integrating the forward and backward computation results, the accuracy of north component of gravity vector improved from 17.71 m Gal to 4.87 m Gal and the accuracy of east component of gravity vector improved from 4.55 m Gal to 4.02 m Gal. Thirdly, the systematic bias and low-frequency trends were corrected by EGM2008. The computed gravity value from EGM2008 reflects the long-wavelength information of the gravity. Although this long-wavelength information can not reflect the high-frequency details of the gravity, they can be used to remove the bias and low-frequency trends in the measurenet results. In Penglai campaign, the accuracy of north component of gravity vector improved from 17.71 m Gal to 2.74 m Gal and the accuracy of east component of gravity vector improved from 4.55 m Gal to 3.20 m Gal after corrected by EGM2008. In Nanhai campaign, the accuracy of north component of gravity vector improved from 50.24 m Gal to7.46 m Gal and the accuracy of east component of gravity vector improved from 48.11 m Gal to 4.48 m Gal after corrected by EGM2008.