Study On Trajectory Planning And Trajectory Tracking Control Of A Meal-Assistance Robot

A meal-assistance robot is a class of service robot that has rarely been studied in China but is indispensable.It is mainly used to assist the disabled or partially disabled elderly and disabled patients to complete their diet.which will not be able to be replaced by the alleviation of the problem of rehabilitation medical services for the elderly and the disabled.Therefore,in order to help the meal-assistance robot accomplish the task smoothly,accurately and efficiently,it needs to make reasonable trajectory planning and choose the appropriate trajectory tracking control algorithm.To this end,we design a 6-DOF meal-assistance robot and use it as an object to research kinematics,dynamics modeling,trajectory planning and the algorithms of trajectory tracking control.Firstly,a 3D model of meal-assistance robot is established by using 3D software and a positive kinematics equation is established by D-H parameter method to solve the attitude matrix of meal-assistance robot,and an inverse kinematics equation is established by analytic method.Secondly,due to the existence of multi-solvability and singularity problems in solved the problem of inverse kinematics by analytic method,a BP neural network based on LM algorithm is proposed to solve the joint variables.In order to pursue higher accuracy,faster convergence speed and minimize the position error in unconventional tasks.a new method based on BP network of traditional LM algorithm is proposed,which adds the feedback method that takes the output of the previous iteration is taken as the next input.Simulation results show that the algorithm can achieve higher accuracy and faster convergence speed.Then,on the basis of the previous kinematics analysis,planning the trajectory of the meal-assistance robot.Mainly using 5 times quasi uniform B spline interpolation method to help the robot trajectory planning.The simulation results show that the proposed method can not only ensure the continuity,smoothness and accuracy of motion curves of the robot's joint displacement,joint velocity,joint acceleration and jerks,but also the planned curves can be well fitted with functions.The trajectory of end effector obtained through positive kinematics has also passed the initial point,taking food point and feeding point of feeding robot task.This indicates that the trajectory of this method can be trusted,and it can complete the task of helping meals while ensuring that food is not spilled.Finally,used the lagrange equation method to build a dynamic model of a meal-assistance robot This model is a mathematical model for trajectory tracking control,and a brief introduction is given to several kinetic characteristics that need to be used.due to traditional algorithms such as PID can not be ensure the accuracy of tracking control,an intelligent control algorithm of the adaptive robust control system based on fuzzy compensation is proposed for trajectory tracking control of meal-assistance robots.Among them,fuzzy control is used to compensate for the uncertainty error of solving dynamic model and the error caused by uncertainty disturbances such as friction,variable load and external disturbance.The simulation results show that the control algorithm has high tracking accuracy,short convergence time and good robustness,and then verifies the algorithm is feasible.