Research On Attitude Synchronization And Force Feedback Design And Algorithm Of Humanoid Manipulator

With the intelligentization of robots and the popularization of applications,humancomputer interaction has become the focus of research in the field of robotics.The data glove with force feedback is an important implementation of human-computer interaction.The data glove developed based on motion capture technology can realize the attitude acquisition of the fingers and arms of the human hand,and convert the collected finger bending angle and arm posture data into robotic drive.The rotation angle of the motor realizes the remote synchronous control of the attitude of the robot;the force feedback device is used to reproduce the interaction between the robot and the environment to the human hand,giving the operator an immersive virtual feeling and realizing the subjective subject The will is combined with the manipulator control,which reduces the control difficulty of the manipulator and improves the control efficiency.This paper first introduces the research status of force feedback data glove and motion capture technology at home and abroad,and puts forward the problem of single force release point of force feedback data glove,and proposes the force feedback structure of double parallelogram,which is the premise of single motor drive.The next implementation achieves equal tension on both knuckles at the same time,thereby increasing the operator's experience of presence.The force feedback system generates a specified pulling force at the finger end by dragging the rigid drawstring by the motor,and places hardware devices such as a motor and a driving device on the palm of the hand,thereby reducing the binding force of the force feedback structure to the finger.According to the different states of human finger movement,a three-ring negative feedback PID controller is designed to realize the precise position control of the motor,so that the rope is collected in time when the human hand makes the stretching motion.The fuzzy PID controller with parameter self-tuning is designed to realize the motor quickly.Sensitive torque control to reproduce the interaction between the robot and the environment in time.The attitude acquisition of the arm and wrist is realized by the nine-axis attitude sensor.The Kalman filter fusion algorithm is combined with the advantages of multiple sensors to achieve accurate attitude acquisition.Since the link structure has a certain motion law,the angle information of the link in the force feedback structure is measured by the angle sensor,thereby indirectly measuring the movement angle of the finger of the human hand.Through the attitude calculation algorithm,the posture information of the arm,wrist and five fingers is converted into the motion information of the 11 motors of the robot,and it is packaged and sent to the remote manipulator through the wireless serial port to drive the movement,so that the robot can follow the human hand movement in real time.The embedded hardware system of this paper is designed with two upper and lower PCB boards according to different functions.The upper and lower two layers are connected by board connectors.The hardware circuits of core functions such as motor drive,attitude calculation and power management are designed on the lower board.All communication interfaces between the system and the outside world are designed on the upper board.When the interaction equipment with the system needs to be replaced later,the upper board can be directly updated,which is beneficial to the secondary development of the system.At the same time,the design of the upper and lower boards is reduced.The space occupied by the PCB board is more in line with the design requirements of miniaturization of wearable devices.