Research On Geomagnetic Vector Measurement Error Compensation Algorithm In Aeromagnetic Geomagnetic Detection

Obtaining accurate geomagnetic vector values is of great significance for space geomagnetic navigation.However,in aerospace magnetics,the attitude angle of a stable flight will still vary within ±10 degrees,resulting in geomagnetic vector measurement errors.Therefore,three compensation algorithms are studied in this paper to compensate the measurement error of the geomagnetic vector.Firstly,this paper analyzes the source of error,describes the internal error of the sensor and the principle of the magnetic interference of the aircraft platform,and constructs the corresponding error compensation model.Aiming at the problem that the correction and compensation are difficult to be separated in practical application,a comprehensive interference field compensation model is proposed,which realizes the simultaneous compensation of the sensor internal error and the interference magnetic field.Secondly,based on the traditional T-L linear model,two compensation algorithms based on total field constraint and three-component constraint are established.Due to the singularity of the matrix,the parameter instability is caused when the least squares estimation is used to solve the compensation parameters.The ridge regression estimation algorithm based on the L curve to solve the ridge parameter is studied to solve the compensation coefficient.The simulation experiments show that,the compensation algorithm based on the total field constraint is effective for the compensation of the total field but the compensation effect for the three-components is not obvious,and the compensation algorithm based on the three-component constraint can compensate the total amount and the three-component.Then,a nonlinear compensation algorithm based on BP neural network is established.The three-component measurement value and the attitude angle of the fluxgate are taken as inputs,and the standard geomagnetic three-component value is output as the IGRF model.Aiming at the problem that BP network initial weight is random and easy to fall into local optimum,a genetic algorithm is introduced to optimize the network.The flight measured data is compensated by the network,and the errors of the compensated three axes are within 4nT,10 nT,and 25 nT,respectively.Finally,a ground pre-experiment platform and a flight test platform for low-altitude geomagnetic exploration were designed.The experimental data was compensated by the compensation algorithm studied in this paper.The results show that the compensation algorithm can significantly improve the geomagnetic vector measurement error.