Research On Head-eye Coordination Motion Control System Of Bionic Robot

With the widespread application of bionic technology in the robot manufacturing field,the combination of visual and auditory applications makes bionic robots more humanized and intelligent.The head-eye coordination motion control system of the bionic robot studied in this thesis includes the motion control system,the auditory system and the vision system.By installing industrial cameras and microphone arrays on a bionic robot to imitate the eyes and ears of creatures,it is possible to obtain the position of the speaker's face through vision and hearing.According to the structure and movement characteristics of the head and eye of primates,a coordinated motion control strategy for each joint is developed.The above features make the system more bionic.In the process of researching on the motion control system,according to the structure of the bionic head-eye coordination motion apparatus and the rotation characteristics of each joint,the system's coordinate model is established,and two industrial cameras are used as the end of the actuator to perform the positive kinematics analysis and inverse kinematics analysis.After the auditory system and the visual system acquire the position of the speaker's face,the motion control system applies head-eye coordination motion control strategy to achieve real-time tracking of dynamic face with both eyes based on kinematics analysis results;In the process of researching on the auditory system,using the AD7606 module to synchronize the acquisition of four-channel signals output from the microphone arrays.The rough spatial position of the speaker's face is calculated by using the generalized cross-correlation time delay estimation algorithm and the three-dimensional coordinates solving method of the sound source.Integrated software development kit provided by IFLYTEK.The speaker can control the running status of the system through voice;In the process of researching on the visual system,using opencv-based Haar+Adaboost face detection algorithm and compressive sensing tracking algorithm to achieve real-time detection and tracking of the speaker's face position.The exact position of the center of the speaker's face is calculated by combining the tracking results with the binocular three-dimensional measurement algorithm of real-time calibration of the external parameters.The experimental results show that the head-eye coordination motion control strategy of the motion control system can achieve real-time tracking of dynamic faces,and maintain the face in the video image recognized by the camera.The head-eye coordination motion control strategy of combining robot kinematics and biomimicry is adopted to expand the robot's perception range and improve the robot's movement flexibility.Based on this,a speech interaction system is added to improve the human-computer interaction characteristics of the system,making the head-eye coordination motion control system of the bionic robot studied in this thesis have a good practical value.