| The detection of magnetic anomalies on aviation platforms refers to the detection of magnetic targets that can cause magnetic interference in the target area through the magnetometer sensors on the aviation platform.The technology originated during the Second World War.Due to its high efficiency,it has now played an important role in geological surveys,unexploded bomb detection,archaeology,etc.In recent years,due to the increasing accuracy of magnetometers,the magnetic interference of aviation platforms has become a bottleneck restricting the detection of magnetic anomalies on aviation platforms.With the continuous development of computer technology,we can accurately establish and solve the model of the magnetic interference of the aviation platform,and use the information such as the movement attitude of the aviation platform to calculate and remove the magnetic interference generated by the aviation platform,so as to continue to judge the existence of the target signal.Traditionally,we generally use the Tolles-Lawson model to describe and compensate.Because the widely used magnetic interference model is too simplified and there are problems such as morbidity and time variability,the compensation effect for the magnetic interference of aircraft platforms is often poor.In our compensation process,the adaptive filter algorithm is used in most cases.Recently,the kernel method has been increasingly used in filter design.At the same time,on the other hand,the target detection algorithm should extract the features of the target and use these features to judge the target.When the kernel method can map data to high-dimensional space,it can more effectively extract target features and find hyperplanes that match the features between their data.Based on the above description,the main work of this article is:Firstly,in view of the time-varying and morbidity of the aviation platform magnetic compensation model,an aviation platform magnetic interference compensation algorithm is proposed.In the aviation platform magnetic interference model(T-L model),the independent variables are the direction cosine of the aviation platform in the coordinate system and their derivatives,so there is a correlation between the independent variables,which leads to the morbidity of the model.In the compensation process of aviation platforms,we will assume that the ambient magnetic field is uniform.But in reality,the magnetic characteristics of the aircraft will change the hysteresis loop,which in turn affects the magnetic field.The algorithm uses the specified window to specify the weight of the iterative data for the time-varying model.At the same time,the objective function is changed during the algorithm iteration to account for the error term to avoid the great influence of the error in the ill-conditioned equations.Test results based on digital simulation and measured data verify the effectiveness of the algorithm.Secondly,a magnetic anomaly target detection method based on kernel adaptive filtering is proposed.This method uses a kernel function to map the magnetic field data to a high-dimensional space,uses the time series modeling method to sort the data,and inputs the algorithm to iterate.At the same time,a feature called "surprise" is constructed.The output of the algorithm is this feature,and the existence of abnormality is judged according to the threshold.The problems of false alarm and low detection rate of the algorithm are analyzed,and then the algorithm is improved.The algorithm can effectively detect abnormal signals.Finally,for the simulation and visualization needs of magnetic anomaly target detection,a visual simulation system for magnetic anomaly detection of aviation platforms was designed and developed.The functions of path planning,data simulation and key compensation,and target recognition algorithm are integrated into the software,which provides an efficient auxiliary tool for the analysis of magnetic anomaly target detection data,path planning and related analysis. |
PDF Full Text Request