Investigations On The Theory And Method Of Magnetic Anomaly Detection

Magnetic anomaly detection technology is widely used in geological exploration,unexploded ordnance removal and submarine detection,and has great value for both civil and military.With the development of UAV technology,magnetic detection system based on small UAV has become a research hotspot.On the one hand,due to the limitation of load and power supply,it is difficult for small UAV to carry the existing high-performance magnetometer,so it is necessary to develop a miniaturized,modular and inexpensive magnetometer.On the other hand,the magnetic field interference of small UAVs is more complex,which puts forward higher requirements for magnetic measurement signal pre-processing and anomaly detection methods.In order to solve the above problems,the influencing factors of magnetic anomaly detection,the time-frequency characteristics of magnetic anomaly signal and the distribution law of geomagnetic noise are deeply analyzed in this paper.On this basis,the theory and method of magnetic anomaly signal detection are studied,which provides theoretical and methodological support for signal pre-processing and anomaly detection of small UAV magnetic detection system.The main work and innovations are as follows:1.According to the characteristics of magnetic anomaly detection of small UAV,the theory and method of magnetic anomaly signal detection,which integrates signal pretreatment and binary hypothesis test,are proposed.The magnetic anomaly detection describes the process of detecting the magnetic anomaly signal produced by the target effectively when the magnetometer is in relative motion with the target(the distance between the two is from far to near and farther),so as to realize the target detection.Magnetic anomaly detection mainly includes two cases: search and monitoring.The former is often used in aeromagnetic exploration.When the magnetic anomaly detection system is fixed on the UAV,the measured magnetic anomaly signal is susceptible to background noise and platform interference,and the magnetic anomaly signal is a finite signal with a non-periodic energy.In this paper,the theory and method of magnetic anomaly signal detection,which integrates signal pretreatment and binary hypothesis test,is proposed.First,the disturbance and error with clear suppression rules are suppressed by signal pretreatment,and then the anomaly signal is detected by binary hypothesis test.2.Aiming at the influence factors of three-axis magnetic sensor error,carrier magnetic field interference and coordinate system misalignment error,the corresponding pretreatment methods are studied.On the one hand,by establishing the mathematical model of error and interference,using least squares and ellipse fitting to calibrate parameters,error correction and interference compensation are realized,the accuracy of total measurement is improved,and the total measurement accuracy of triaxial magnetic sensor reaches 1nT.On the other hand,a comprehensive error model of Strapdown geomagnetic vector measurement system is established,and a set of equations constrained by "vector+scalar" is constructed.The Lagrange multiplier method is used to accurately estimate the integrated calibration parameters,thus the measurement errors can be calibrated together,which can effectively improve the accuracy of geomagnetic vector measurement.3.Aiming at the problem of magnetic anomaly detection in search mode,an orthogonal basis function(OBF)correlation detection method based on Kahnan-Lowe expansion(KLE)is proposed,which effectively overcomes the influence of colored noise.In search mode,in order to reduce the effect of attitude jitter,magnetic scalar is often used as effective information.In this paper,based on the Kahnan-Lowe expansion principle and the autocorrelation matrix of geomagnetic noise,the standard orthogonal basis function is modified,and the KLE-OBF detector is constructed.Optimal detection under the background of color noise is realized,and the parameters of the detector are optimized.The results of simulation and experiment show that under the premise of false alarm probability 1.5% and signal-to-noise ratio-3d B,the detection probability of KLE-OBF detector is increased from 70% to about 80%,compared with standard OBF detector.4.To solve the problem of magnetic anomaly detection in monitoring mode,a vector correlation detection method is proposed,which further improves the detection probability of magnetic anomaly signal.Based on the magnetic dipole model,the orthogonal basis function of the magnetic anomaly vector is derived,and the vector correlation detector is constructed.A frequency-domain adaptive denoising method is proposed for dual-sensor monitoring system,which can effectively suppress common-mode geomagnetic noise and interference.This method makes full use of geomagnetic reference information and three-component measurement signal,so it can effectively improve the detection probability of magnetic anomaly signal.Moreover,with the continuous improvement of measurement accuracy of Strapdown geomagnetic vector measurement system,this method can also be applied to search mode.The simulation and experimental results show that the detection probability of vector detection is increased from 80% to about 97% under the premise of false alarm probability 1.5% and signal-to-noise ratio-3dB,compared with OBF detection based on scalar information.5.In view of the absence of prior information of signals,a method based on stochastic resonance for detecting magnetic anomaly signals is proposed,which guarantees robust detection performance.In the absence of prior information of signals,it is difficult to detect the magnetic anomaly signal effectively by the general non-parametric detection method under low signal-to-noise ratio(SNR).Stochastic resonance(SR)system can improve SNR by converting noise energy to signal,thus ensuring the effective detection of the magnetic anomaly signal.Firstly,the signal enhancement mechanism of SR system and its numerical simulation method are studied;secondly,according to the time-frequency characteristics of SR output signal,a detector based on parallel SR system(PSR detector)is designed,namely,the measurement is enhanced by PSR system firstly,and then is detected by an energy detector.The experimental results show that under the premise of false alarm probability 1.5% and signal-to-noise ratio-3dB,the detection probability of PSR detector can be stabilized about 70%,which is obviously better than that of minimum entropy detector(about 55%).6.The UAV magnetic anomaly detection system is built and the magnetic anomaly detection experiment is carried out to verify the effectiveness of the proposed method.Based on a small three-axis fluxgate sensor and a rotorcraft UAV,a UAV magnetic anomaly detection system was built.The validation experiments of UAV interference compensation,flight search and hovering monitoring were carried out.After interference compensation,the total fluctuation of magnetic field caused by UAV attitude change is reduced from 78.7nT to 3nT,the compensation improvement is comparable to existing aeromagnetic compensator's.The experimental results show that the KLE-OBF detector and the PSR detector can still detect the magnetic anomaly signal effectively under the premise that the signal-to-noise ratio of the measured magnetic anomaly signal is as low as-3dB.Generally speaking,this paper demonstrates the feasibility of magnetic target detection by the fixed magnetic anomaly detection system of the small-scale rotorcraft UAV,and realizes the detection of magnetic anomaly signal in the background of strong interference/noise.