Underwater Target Positioning Technology Based On Magnetic Gradient Tensor

The detection of underwater target based on the magnetic anomaly is promising and widely used. With the influence of geomagnetic field, ferrous objects will produce magnetic anomaly signal and disturb the distribution of geomagnetic field around those objects. By detecting and processing magnetic anomaly signal, the location information of magnetic targets can be calculated. This paper do some research on the target positioning technology which is based on magnetic gradient tensor:Firstly, some research is done on the analytic signal method of the magnetic gradient tensor. There are three kinds of calculation formula for the analytic signal of the magnetic gradient tensor, and they are Ax, Ay, and Az. The Location results of the three formulas are analyzed. Some influence factors of results are discussed, including the magnetic inclination,magnetic declination and noise.Secondly,the magnetic gradient values of the measurement point are needed for magnetic gradient tensor method. If the noise is strong and the signal is weak, the magnetic gradient value can't reflect the gradient feature of signal obviously. Savitzky-Golay filtering method is commonly used in differential filter, and it's applied to the calculation of magnetic gradient tensor. The simulation results show that Savitzky-Golay filter can reduce the noise influence effectively, and improve the accuracy and astringency of location results. This paper also studies the choice of the filter parameters of Savitzky-Golay filter.Thirdly, the analytic signal method is simple and the precision single point method is high, the joint inversion method that based on magnetic gradient tensor is put forward. In the process of joint inversion, the analytic signal method is used to get the preliminary horizontal position of target. The single point method completes the final location with the magnetic anomaly the magnetic gradient tensor information of the preliminary point. The simulation results show that the positioning accuracy of the joint inversion method is higher than the analytic signal method.