Design Of Unattended Acoustic Shock Detection Device

Unattended target detection system is widely used in military target reconnaissance,identification and intrusion protection around special buildings.Aiming at the problems of low target recognition accuracy,low positioning accuracy and limited detection distance in practical application,this paper carries out the design of unattended acoustic vibration detection device based on the in-depth analysis of the characteristics of typical target ground motion signals and acoustic signals.By analyzing the frequency characteristics and attenuation characteristics of typical target seismic waves and sound waves,the combined measurement method of target recognition and sound location based on seismic ground motion is determined.The performance of several kinds of ground motion sensors was studied experimentally and the sensor based on magnetic induction principle was selected as the target ground motion signal measurement sensor.According to the combined sensor performance parameters and output signal characteristics,the preamplifier method is adopted to improve the gain of the measurement front end,and the detection range reaches 50 m.EMD de-noising method are analyzed with wavelet threshold denoising method in seismic signal denoising processing characteristics,in view of the low frequency characteristics of seismic signals,this paper presents an improved EMD-wavelet threshold denoising method,the experimental results show that the method can effectively suppress the noise in the seismic signal and can keep the effective characteristics of a target.Analyzing and extracting the eigenvectors in the ground motion signals,using the genetic algorithm to optimize the BP neural network model and to identify the human and the car,and the results show that the method has a better convergence than the traditional BP neural network,and has a higher rate of recognition and calculating speed.In view of that frequency characteristic of acoustic wave signal of typical target in the measurement environment,the propagation attenuation law of acoustic wave in air is studied,and the sound position method based on TDOA positioning principle is established.Systematically analyzed the actual working conditions is unattended target positioning error factor of detection system,wind speed,sensor installation error and interpretation due to the sensor consistency,regulate circuit and feature points caused by random error factors such as time error,azimuth calculation error of the simulation analysis of the positioning model,determine the orientation accuracy allowed within the scope of the applicable environment and positioning,verify the reliability of the positioning model.An unattended acoustic vibration detection device is designed.Multistage amplifier,filter circuit and voltage comparator circuit are adopted to realize multistage amplification,filtering and de-noising of the output signal of the sound sensor and shaping processing;Based on FPGA and STM32,the time difference extraction module of acoustic signal and the acquisition module of ground motion signal were designed to realize the acquisition,storage and transmission of target signals;Based on LabVIEW,the upper computer software is designed to realize the realtime control and target display functions of the system.By simulating the unattended environment in the field,the experimental research of the device is carried out,and the target positioning and recognition function of the unattended device is verified.The positioning error and control method are analyzed;Experimental research shows that,within the scope of 30 m,personnel and vehicle positioning accuracy is within 5 °,the recognition rate is 98.96%.