Research Of The Robotic Arm Teleoperation Technologies Based On Gesture Recognition Using Inertial Sensors

With the development of industrial automation and intelligence,the robotic arm is taking on more and more production tasks.Traditional robotic arms require advance programming to control,resulting in a lack of flexibility and intelligence in their operation.Therefore,humanrobot collaboration has become a development direction for achieving greater intelligence in robots.The use of teleoperation for human-robot collaboration allows humans to help robots perceive and understand the environment,improve robot performance,and efficiently complete complex tasks.In this study,the core idea of gesture teleoperation of the robotic arm is to convert human intent through gestures into motion commands for the robotic arm,achieving control of the robotic arm.A gesture recognition-based robotic arm teleoperation system using inertial sensors was designed.The main research focused on human body posture detection and tracking based on inertial sensors,static gesture recognition methods in teleoperation,dynamic gesture mapping methods,and safety evaluation methods in human-robot collaboration.The specific research contents are as follows:(1)The inertial sensors are used to track the human body posture.The extended Kalman filter algorithm is employed to fuse the data from the inertial sensors and to perform attitude estimation.Quaternions and Euler angles are used to describe the attitude of the inertial sensors.By simplifying the human body structure,a kinematic analysis of the human body is conducted using a chained skeleton model to achieve real-time tracking of the human body posture using inertial sensors.(2)Static gesture recognition and dynamic gesture mapping methods are designed to implement a fusion of static and dynamic gestures for teleoperation in the teleoperation system.The arm posture data obtained from the inertial sensors are used as the control input for the teleoperation system.The wavelet denoising method is used to filter the raw data used in gesture recognition,improving the accuracy of gesture recognition.An analysis is conducted to determine the appropriate use of static gestures in the teleoperation system,and a gesture recognition method based on arm vector features is proposed.The motion of the human arm is analyzed to obtain the required end-point posture of the human arm in dynamic gesture control,and the dynamic gesture master-slave mapping is implemented using an incremental proportion mapping method based on the workspaces of the robotic arm and the human arm.(3)Based on the constructed teleoperation platform,a teleoperation system integrating gesture recognition and robotic arm control functions was established,and teleoperation system software was developed.By tracking the position and orientation of the robotic arm and the human body,the distance between the human and the robot was detected,and a safety evaluation method was proposed by combining the running speed of the robotic arm to ensure the safety of human-robot interaction within the same workspace during teleoperation.Experiments verified the effectiveness of the gesture recognition and master-slave mapping methods as well as the human-robot safety evaluation method of the teleoperation system.A teleoperation experiment in a simulated real-world scenario demonstrated the flexibility and efficiency of the teleoperation system in controlling the robotic arm to complete operational tasks.In summary,this study developed a mechanical arm remote control system based on gesture recognition using inertial sensors.The gesture recognition is based on the data obtained from the inertial sensors and serves as the input for the teleoperation system.The system integrates dynamic gesture mapping and static gesture instruction to control the mechanical arm,providing a natural interactive experience,high safety,and efficient remote operation performance.