| The seventh national census data shows that the number and proportion of China’s elderly population has increased significantly compared to a decade ago,and the speed of aging is accelerating,and the contradiction between the demand for assisted feeding of the elderly with upper limb disability and the insufficient number of nursing staff is becoming increasingly prominent.In order to solve the daily dining problems of the elderly with disabled upper limbs,the research and development of meal-assistance robots will focus on the problems of low intelligence,simple motion control process and single human-computer interaction mode of current meal-assistance robot products.This paper analyzes the application scenarios and service targets of meal-assistance robot,formulates robot design indicators,and designs robot structure schemes,control schemes,and human-robot interaction schemes.The robot structure consists of three parts:rotating dinner plate,food picking robotic arm,and spoon and chopstick integrated hand;The control system adopts a hierarchical control scheme,which is composed of two main control systems:the upper computer control system and the lower computer control system;In order to adapt to the meal assistance needs of people with different degrees of upper limb disabilities,the humancomputer interaction mode adopts a multi-modal human-computer interaction scheme that combines mechanical interaction and natural human-computer interaction.To improve the intelligence level of the meal assistance robot,realize the robot automatically picking up and delivering meals,meal assistance robot needs to build a visual recognition system and establish the face-mouth detection and positioning model to locate the target location of meal delivery accurately and provide data support for the motion planning and visual servo control during meal delivery.This thesis uses the D-H method to establish the robot kinematic model and the analytical method to solve the robot inverse kinematics;analyzes the robot motion characteristics,divides the robot into five motion phases,and focuses on planning the motion trajectory when the robot picks up and delivers the food.The robot pickup position parameters can be obtained by teaching the robot during the pickup.The meal pickup trajectory can be planned by using Bsample curve,and based on this,an optimized trajectory over critical path points can be formed by using INSGA-II algorithm with time,energy and impact as the objectives;The robot vision system can locate the target position in real time,and adopt a multi-segment linear interpolation based on seven-segment S-shaped acceleration or deceleration planning and posture interpolation model based on unit quaternions are used to plan the delivery trajectory for safe and accurate food delivery.According to the needs of robot control,Altium Designer is used to design robot motion control and power management hardware control circuits based on ARM microcontrollers and power management chips;The robot control also needs to analyze the robot task execution process,develop the robot up and down network communication protocol,and use QT to design the robot upper computer operation interface;According to the design of the robot hierarchical control scheme,the underlying driver of the lower computer and the robot motion control logic timing program are built using Stm32CubeMx and MATLAB/Stateflow respectively.On the basis of hardware and software control system development,we design robot vision servo control strategy based on face position tracking,and establish robot whole machine PID control and motor closed-loop control algorithm model.We build the engineering prototype of the meal-assistance robot,test the hardware and software performance of the control system,test the basic functions of the robot such as meal selection,meal picking and meal delivery,and analyze the rationality of the planned traj ectory,the realization of each basic function and the servo tracking performance of the robot. |
PDF Full Text Request