Research On Self-gradient Correction Method Of Aviation Gravity

In airborne gravity gradient measurement,the attitude change of the carrier produces self-gradient effect on the gravity gradient meter with ultra-high sensitivity,which is a serious interference to the measurement results.Because of the complexity of aircraft structure,it is difficult to remove this kind of interference precisely by simple mass modeling and forward modeling.In view of the shortcomings of the existing methods,a data-driven self-gradient effect correction method is proposed based on the basic theory of supervised learning.The purpose is to reveal the relationship between attitude change and self-gradient effect of the carrier and lay a foundation for self-gradient correction of airborne gravity gradient measurement.The main contents of this paper include:Functional approximation is used to study the relationship between self-gradient and aircraft attitude.Based on Fourier series fitting and polynomial fitting,three self-gradient functions of attitude angle and self-gradient relationship are obtained by using the mathematic solution strategy of high-order nested regression analysis.The simulation training experiment is conducted to set different flight attitude and rotate the aircraft model to get the self-gradient value.On this basis,white noise is added and divided into two parts: one part is used as training set,and the corresponding coefficients are obtained by using Fourier series and polynomial fitting method respectively;the other part is used as verification set,and the order is adjusted continuously with the objective of minimizing the error of verification set.The optimum self-gradient function is obtained.The simulation flight experiment is carried out,the underground anomalous body is set up,the gravity gradient is measured in the observation plane,and the self-gradient interference caused by the random change of aircraft attitude is added.The optimum function and coefficient obtained from the simulation training experiment are used to calibrate the self-gradient of all the signals of the full tensor gravity gradient measurement.The results show that the real data generated by the anomalous body and the corrected data are basically maintained.It holds the same order of magnitude and preserves the anomalous shape completely.It can correct all the signals of full tensor gravity gradient measurement.Based on the above research contents,the application software of self-gradient correction for airborne gravity gradient measurement is designed.The simulation calculation of different attitude self-gradient,the calculation of coefficient of self-gradient function and the calculation of self-gradient correction are realized in the process of airborne gravity gradient measurement,and the correction results are displayed graphically.