Design And Realization Of Space Target Location Of Artificial Bat-like Ears

As a "live radar" in nature,bats can use their wonderful sonar system to accurately locate targets in dark and complex environments,and are far superior to current artificial sonar in terms of accuracy and sensitivity.Numerous studies have shown that the shape of bat ears is related to this ability.However,in the process of imitating the bat sonar system,researchers often ignore this factor.Plecoutus auritus are widely distributed in China,with huge ears as their typical feature and a super sonar system This article summarizes previous experience,by strictly imitating the ear structure of the Plecoutus auritus,and imitating the active sonar system of the bat,a set of artificial bat ear space target positioning methods are proposed,And through bat ear space target positioning signal acquisition platform to achieve multi-index accurate positioning of the target in space.Firstly,this article divides the target positioning process into three parts:determines whether the target exists,determines the distance between the target and the acoustic signal receiving means and the DOA estimation of the target.The method of preprocessing the acoustic wave signal is introduced.The feasibility of determining whether the target exists through the cross-correlation threshold method is introduced The feasibility of the autocorrelation analysis method and the convolutional neural network distance classification method to determine the target distance is introduced The target DOA estimation method using the combination of local linear embedding algorithm and BP neural network is introduced.Secondly,this paper obtains the Plecoutus auritus ear model through 3D printing technology,and builds on this model to build a bat-like ear space target positioning signal acquisition platform.The signal acquisition platform imitates the process of bats performing active sonar,and emits ultrasound waves by imitating the mouth of the bats through the ultrasound player,and the bat ear model and the ultrasonic microphone built in the root of the bat ear model form an acoustic wave signal receiving device,which receives the acoustic wave signal reflected back by the target,and by controlling the rotation of the stepping motor placed at the bottom of the sound wave signal receiving device,the process of imitating the rotation of the bat ear is realized.The sound wave signal collected by the acoustic wave signal receiving device is converted into an audio file through the NI PXI control system.Finally,relying on the bat-like ear space target positioning signal acquisition platform built for a large number of experiments.The endpoint detection method is used to intercept the sound wave signals other than the emitted wave from the binaural sound wave signal,and the cross-correlation threshold method is used for the binaural sound wave signal to determine whether the target exists,and the judgment accuracy rate is above 90%.Classify the acoustic signal spectrum comprising a transmission wave echo by a convolution neural network,determines the horizontal distance between the target and the acoustic signal receiving means,and the judgment result is not affected by the pitch and azimuth of the target.Using echo interception and signal feature extraction methods to extract features from the acoustic wave signal,and using the local linear embedding algorithm to reduce the dimension of the extracted features,and dimensionality-reduced features are used as input,and the DOA estimation of the target is performed through the BP neural network.The accuracy of determining the pitch angle is high.When the target is above the horizontal plane where the root of the bat ear model is located,the accuracy of determining the azimuth angle is high.The results show that the proposed artificial bat ear spatial target positioning method can achieve the requirements of positioning the target.It provides a new solution for the current robot to use sonar to locate the target.