Research On Intrusion Detection Technology Of Offshore Coast Based On Unattended Ground Sensor

With the development of social economy,national defense and military,our nation has continued to explore the resources of coastal ecological in recent years.The increasing number of coastal military and civilian key facilities and regional security issues have attracted people's attention.The traditional coastal protection measures are mainly realized by combining traditional technical means such as buoys,photoelectric warning radars,sonar,and physical means,whose focus of protection is only on water surface and underwater protection.However,the coastal protection as the last protective barrier near the coast has not received widespread attention.UGS(Unattended Ground Sensor)system is an unattended integrated detection system that can independently realize monitoring,detection,and measurement functions.Compared with the widely used video surveillance,infrared detection,sound-controlled alarm and other facilities and equipment in the security field,UGS has better concealment,full-time,applicability and stability in the field of coast protection.Therefore,the technology of UGS-based near-coast border intrusion detection can become the last defensive line for offshore security,and it also can be a complement of existing technologies and equipment in the offshore security field.In the special coastal environment,the separated detection technologies such as,image,sound,magnetic,and vibration signals cannot fully meet the application requirements due to their technical limitations.Among them,the vibration signal detection technology is less affected by the coastal environment.This thesis conduct the research on the basis of UGS(Unattended Ground Sensor)system,taking this as an entry point for in-depth research and mainly focusing on five aspects which included the characteristics analysis of seismic signal,the acquisition system design of seismic signal,the feature extraction of seismic signal,the detection model construction and verification of seismic signal,and the analysis of the impact of near-coast environments on seismic signals.Firstly,the generation and propagation mechanism of the near-coast vibration signal is studied and the waveform composition of the vibration signal,the attenuation of each waveform and the proportion of energy are discussed.Selecting the Rayleigh wave as the research object of the vibration signal,this thesis emphasized the analysis of the characteristics near-coast vibration source,propagation medium characteristics and acquisition characteristics of the signal.The acquisition of high-quality vibration signals requires good acquisition equipment.In view of the characteristics of near-shore vibration signal acquisition,this thesis firstly selects a passive moving-coil vibration sensor which is with small size,balanced performance and stable structure,and then designs a low-noise amplification and conditioning circuit.At the same time,use anti-corrosive materials to design a waterproof collection node housing and perform leaching.The water and immersion experiments also are applied to justify the results.Finally,the vibration signal acquisition software was written by using Lab VIEW and MATLAB mixed programming,which was loaded on a laptop as a signal acquisition terminal,and used with NI's USB 6218 acquisition card to realize the overall design of the vibration signal acquisition system.Meanwhile,according to the requirements of seismic signal characteristic analysis and feature extraction,a large number of seismic signals under various conditions have been collected,which provides substantial data accumulation for the study of near-coast seismic signals.The feature extraction process of near-coast ground vibration signals is divided into three steps: signal preprocessing,feature analysis and extraction,and feature vector construction.Firstly,this thesis selects the optimal wavelet threshold denoising as the signal preprocessing algorithm through simulation comparison,and then analyzes the characteristics of ground motion signals from three aspects which are time domain,frequency domain,and time-frequency domain,and by using feature extraction procedures to extract variance,peak-to-peak,and the average square root,total energy in frequency domain,energy of different sub-bands,wavelet energy entropy and singular entropy as the characteristics of ground motion signals under the conditions of three types of propagation media of fine sand,coarse sand and gravel.Finally,three types of propagation media are constructed according to the extraction results eigenvectors of near-coast ground motion signals under different conditions.Finally,a two-class optimization logistic regression algorithm based on machine learning-Logistic regression model is used to detect the noise of the coastal environment and the vibration signals of human intrusion.Three vibration signal detection models of fine sand,coarse sand and gravel are built and a large number of field collections are used.The data is verified,and the detection accuracy rates are respectively above 90.75%,82%,and 77%.The fine sand and coarse sand detection models meet the application requirements,and the gravel detection model needs to supplement the training set for regression coefficient optimization.At the same time,according to the results of model verification and signal analysis,the impact of six types of environmental factors on the vibration signal is analyzed for the periodicity of the nearshore wave,the intensity of the nearshore wave,the type of propagation medium,the temperature and humidity of the propagation medium and the deployment method of the collection node.It laid the foundation for the practical application of UGS seismic signal detection technology near the coast.The research direction that combines the two fields of border security and marine and coastal research has huge research space and application value.The research on near-coast border intrusion detection technology can provide more important military and civilian facilities on the coast,uninhabited islands,and overseas bases.In order to improve the three-dimensional monitoring and defense system against the sea,it is of great significance to safeguard the safety of people's property and defend national sovereignty.