Exploration Of Magnetic Anomaly Detection Methods Under Geomagnetic Background

The existence of the geomagnetic field makes the ferromagnetic material be magnetized,thus producing the magnetic field,which causes the local magnetic field to change abnormally.The distribution of the new magnetic field generated by the magnetic target that changes the magnetic field of the surrounding environment is called magnetic anomaly.This newly generated magnetic field is a prerequisite for magnetic anomaly detection(Magnetic Anomaly Detection,MAD).MAD is a passive method for detecting ferromagnetic objects.It has been applied in many fields,such as unexploded ordnance detection,underwater pipeline or cable detection,human medical examination,aerial magnetic detection,metal mineral detection and earthquake Forecast.Since the size of the magnetic field decreases rapidly as the distance increases,the magnetic anomalies generated by remote targets are usually buried in magnetic noise,therefore,effective methods are needed to improve the performance of magnetic anomaly detection.People use the high sensitivity magnetometer to measure the magnetic field containing the target,through the analysis of the measured data,the position of the measured target,volume,and then realize the idea of detecting the target,to achieve the goal of target identification and positioning.In this way,the idea of detecting the target is realized,and the purpose of identifying and positioning the target is realized.This article explores the detection methods of magnetic anomaly signals from the following aspects:First,this paper introduces the construction principle of the three magnetic anomaly detection methods in detail.One is the four-standard orthogonal basis function detection algorithm using the characteristics of the target signal.This method detects the target by matching the orthogonal basis detection and solving the normalized energy function of the target.The second is a differential entropy filtering method based on noise background,which uses the probability of collecting magnetic field data samples to calculate the entropy value.The existence of magnetic anomalies is revealed by the entropy value falling below the threshold.The third is the joint detection algorithm.Based on the target and noise are two different detection methods,each has its own advantages and disadvantages.Based on the four-standard orthogonal basis function decomposition algorithm and the differential entropy filtering detection method,in this paper,a joint detection algorithm is proposed,which more effectively combines target and background features to achieve the detection of target signals.Second,the actual detection of targets is mostly in an unenclosed natural environment,so the surrounding magnetic field noise is the top problem encountered in the signal acquisition process.In the literature on magnetic anomaly detection methods,the main component of geomagnetic noise is generally regarded as colored noise with power spectral density of 1/f~?(0???2).Therefore,we need to analyze the surrounding magnetic field noise,use the AR power spectrum model to whiten the colored noise,and for non-Gaussian white noise,we propose a denoising method combined with wavelet entropy theory to make a better lead for more accurate detection.Third,Biot Savart's theorem is firstly used to theoretically derive the magnetic dipole model,because in the measurement process,the target is generally equivalent to a magnetic dipole for calculation.Then three kinds of magnetic anomaly detection algorithms are used to analyze the simulated magnetic anomaly signals.The results show that the joint detection algorithm has better detection effect.At the same time,multiple magnetic anomaly target signals,which are equivalent to multiple dipoles are analyzed and detected,and the results show that the proposed algorithm has higher anti-leak and anti-error detection rates.Finally,the main technical indicators of the optical pump magnetometer used in this paper are measured,and then the feasibility of the detection algorithm is explored by many experiments in and out of the room.One target and several target Settings were measured in the outdoor experiment.Combined with the actual measurement experimental data,it is proved that the algorithm of joint detection is more excellent and the effect of noise suppression is more obvious.The results show that the joint detection algorithm not only has a good detection effect in the simulation experiment,but also has a high practicability in the actual experiment.