Font Size: a A A

Research On Human-computer Interaction Intention Perception Based On Tactile Information

Posted on:2021-03-12Degree:MasterType:Thesis
Country:ChinaCandidate:Y X LiuFull Text:PDF
GTID:2518306557998989Subject:Mechanical engineering
Abstract/Summary:PDF Full Text Request
With the continuous development and updating of intelligent robots,collaborative robots have developed rapidly and become the hotspot of research today.Robots can help humans to complete simple and repetitive tasks efficiently.They are widely used in military,aerospace,rehabilitation,and industrial fields.However,in uncertain environments or accidents,robot decision-making ability declines and cannot be completely It is an important issue in today's research to replace the human task and give the robot a certain ability to perceive the efficiency and flexibility of human-computer collaboration.However,in contact human-computer cooperation,influenced by factors such as the immature research of flexible array skin sensors and the variable impedance parameters of collaborators,the intentional perception of human-machine collaboration has become a key factor restricting the operability of human-machine interaction.Therefore,in view of the common problem of intent perception and human-computer interaction flexibility in current contact human-machine collaboration,this paper independently designs an array-type tactile sensor that can sense the hand-grip posture and force,and will have 28×28 A perceptual point tactile sensor is packaged on a cylindrical handle,and a convolutional neural network(CNN)-based three-mode state that distinguishes the hand from the grip,the tight grip,and the inadvertent touch on the tactile handle is proposed.The method has a recognition accuracy of 98.2%.Based on this haptic handle,the local hand gesture can be sensed in real time,the operator's operation intention is accurately estimated,and the local perception information is transmitted to the robot to control its motion.This paper takes the UR cooperative robot as the experimental platform,and uses the tactile handle as the perceptual input and carries on the person.The machine interaction experiment evaluated the motion accuracy of the robot arm.Experiments have shown that the tactile handle has good intentional perception.The main research contents of this paper are:(1)Research on the status quo and analysis of contact human-machine collaboration and the design of sensory system model: Based on the research background of human-machine collaboration,we are familiar with and analyze the hotspots of the most advanced contact human-computer collaboration research in today's world.Solve the problem of intent understanding of contact human-computer cooperation by means of multi-information interaction,and complete the mathematical model design of contact human-computer cooperation adopted in this topic.(2)Research on mechanism of operation intention understanding based on touch/force perception information: Design an array type tactile sensor and package it into a tactile handle that can sense the posture and force of the human hand.Based on the haptic handle,the operator perceives the intention information,performs pre-processing on the collected data,and comprehensively describes and defines the operation intention of the contact human-machine cooperation operator.(3)Research on contact man-machine cooperative compliance control method based on touch/force perception: According to the description and definition of operation intention made by research content 2,a compliance control method based on touch/force perception is proposed.The above control signal reflecting the operator's operation intention is input into the robot control system,and the robot motion is driven to respond to the operator's operation intention,and an experimental platform is set up to verify the effectiveness of the method.
Keywords/Search Tags:Human-robot collaboration, Tactile perception, Human grip recognition, Intentional understanding, Compliant control
PDF Full Text Request
Related items