Strict Imitation Space Target Location Of Brown Long-Eared Bat

Using sonar localization in space is an effective supplement to robot visual perception.Most bats can navigate or hunt insects by using sonar,which enables them to move around in dark and complex environment and locate targets.Many studies have linked the shape of bats'ear to ability of localization,but few researchers have developed artificial ears to mimic bat's ability,because it's difficult to determine the target elevation.In this article,we studied the spatial acoustic field characteristics generated by specific structures of the brown long-eared bat(Plecotus auritus),studied the accurate elevation echolocation of bat's ear,and proposed an effective 3D target location method.In addition,we strictly imitate the positioning function of Plecotus auritus.and design a binaural active echolocation system,which can estimate the target angle using active sonar.Firstly,through the finite element simulation calculation of the bat's auricle model at a certain frequency,we made statistics on the simulation results,and determined the basic performance of the frequency-driven beam scanning of the bat ear model.Secondly,in order to verify the positioning characteristics of bat ears in the simulation,a two-ear sound signal acquisition and positioning system is designed and built.In order to control this system to achieve different functions,we design a software and hardware environment to drive it.The artificial bat ear made by 3D printer is the core of the system.and the ultrasonic loudspeaker is the sound device.The ultrasonic microphone is inserted into the ear model as the signal receiver.The rotation of the stepper motor is used to simulate the rotation of bat head.NI signal acquisition card was used for high-speed dual-channel synchronous signal acquisition,and the sound waves collected by the system were saved as audio files.Finally,the target echolocation function in 3D space is realized by using the ear model of Plecotus auritus.We used the acquisition system to carry out the bat ear passive signal acquisition experiment,and processed the collected two-channel sound signals,verifying the bat ear beam scanning obtained by simulation.At the same time,the bat ear beam has a certain resolution at each elevation,and the frequency responsehas formed a certain mode.We conducted the active echo signal location experiment.The target elevation can be determined by different echo signal mode.Using this method,we design and train the neural network.We completed the elevation estimation in parallel binaural and proposed a sliding window averaging algorithm based on pulse train signals to optimize the positioning.In addition,the right angle auricle structure is designed to realize three-dimensional positioning in space.The experimental results show that the bat-like ear localization method used in this paper has higher positioning accuracy.It is also proved that the topolocgical structure of both ears plays an important role in bat echolocation.The conclusion of the study can make people understand the positioning principle of bats profoundly.and the designed echo positioning provides a new method for the localization and navigation of robots in reality.