Research And Design Of The Teaching System For Industrial Manipulator Based On Gesture Recognition

With the rapid development of industry,the requirement of the industrial manipulator is becoming higher and higher.The industrial manipulator should not only accomplish the repetitive work steadily,but also have the characteristics of intelligence,networking,openness,man-machine friendliness.As far as we know,the traditional manipulator teaching system has many disadvantages,e.g.,it is complex to operate and requires professional operator.Besides,the human-computer interaction is not natural.Therefore,it is very important to improve the intelligence of the teaching system and the working efficiency of the industrial manipulator.In this paper,an industrial manipulator teaching system is designed based on gesture recognition,which is intelligent,interactive,natural and efficient.The teaching system mainly consists of three modules.There are sensor module,remote client module and working mechanical arm module.The sensor module is consist of the MYO armband,the MYO armband include 8-channel surface electromyographic(sEMG)sensors and inertial measurement unit(IMU).The IMU is used to collect the position and the posture data from the operator's arm.The sEMG sensors are used to read electrical signals from muscles in arm,which can map the signals to gesture made from hand.The remote client module is mainly composed of data processing module and simulation module.Kalman filtering is used to process the data collected by the sensors,and then convert it to the angle information of the joint angle and gesture instruction.The simulation module is designed based on the ROS,which can show the gesture of the operator's arm in real time.The working arm module is the UR5 manipulator and manipulator working box.In the ROS compiler environment,the remote clients are developed based on Python language development.They are connected with the MYO armband via Bluetooth 4.0 and are connected with the working arm through the TCP/IP protocol.In the way,the teaching system can be demonstrated by directly controlling the manipulator through gesture recognition.The designed teaching system of industrial manipulator is applied into the actual teaching environment of industrial manipulator.The experimental results verify the following functions,firstly,the gesture can control the manipulator directly,secondly,the robot can achieve consistent with the operator's arm.Experimental results show that the design and implementation of gesture recognition based industrial manipulator teaching system can realize the basic functions of the teaching system,which can meet the indicators.It has many advantages,such as high recognition accuracy,friendly man-machine interaction,high intelligence and convenient remote control.