Hello again!
How to do trajectory planning via MATLAB?
What is the best way to do this?
I have computed by hand the forward kinematics via D-H, also the forward differential kinematics (Jacobian matrix) and inverse differential kinematics, but I'm not sure if inverse kinematics is correct.
So I don't want to use inverse kinematics for trajectory planning.
Thanks in advance!
Matlab has an inbuilt inverse kinematics solver function that you can use and can also specify the algorithm that you want to use to solve. You just need to model your system as a RigidBodyTree model
Check out:
For example here is a sample of the creation of a RigidBody joint for the MCP joint on an index finger that is attached to a body called "hand"
%Index MCP joint
body = robotics.RigidBody('IndexL1');
%Indicate as reolutionary joint
joint = robotics.Joint('IndexMCP', 'revolute');
%rotation limits
joint.PositionLimits = [0 pi/2];
%Set location
setFixedTransform(joint,trvec2tform([0 0 0]));
%rotation axis
joint.JointAxis = [0 1 0];
body.Joint = joint;
%connect to index knuckle
addBody(hand, body, 'base2Index');I'm going to check it! Thank you very much!
I tried it and manage to create rigidbodytree and find inverse kinematics, but I don't know how to do trajectory planning.
This is my code so far
%% *** Robot (kinematic) model parameters for Kuka 6-dof ***
clear all; close all; clc;
% Lengths of robot links in mm
l0 = 810;
l1 = 200;
l2 = 600;
l3 = 30;
l4 = 140;
l5 = 550;
l6 = 100;
l7 = 100;
%% D-H parameters
% columns: a, alpha, d, theta
dhparams = [l1 -pi/2 l0 0;
l2 0 l3 -pi/2;
l4 -pi/2 0 0;
0 pi/2 l5 0;
0 -pi/2 l6 0;
0 0 l7 0];
%% Create RigidBodyTree for Kuka
Kuka = rigidBodyTree;
body1 = rigidBody('body1');
q1 = rigidBodyJoint('q1','revolute');
setFixedTransform(q1,dhparams(1,:),'dh');
body1.Joint = q1;
addBody(Kuka,body1,'base');
body2 = rigidBody('body2');
q2 = rigidBodyJoint('q2','revolute');
body3 = rigidBody('body3');
q3 = rigidBodyJoint('q3','revolute');
body4 = rigidBody('body4');
q4 = rigidBodyJoint('q4','revolute');
body5 = rigidBody('body5');
q5 = rigidBodyJoint('q5','revolute');
end_effector = rigidBody('end_effector');
q6 = rigidBodyJoint('q6','revolute');
setFixedTransform(q2,dhparams(2,:),'dh');
setFixedTransform(q3,dhparams(3,:),'dh');
setFixedTransform(q4,dhparams(4,:),'dh');
setFixedTransform(q5,dhparams(5,:),'dh');
setFixedTransform(q6,dhparams(6,:),'dh');
body2.Joint = q2;
body3.Joint = q3;
body4.Joint = q4;
body5.Joint = q5;
end_effector.Joint = q6;
addBody(Kuka,body2,'body1');
addBody(Kuka,body3,'body2');
addBody(Kuka,body4,'body3');
addBody(Kuka,body5,'body4');
addBody(Kuka,end_effector,'body5');
showdetails(Kuka);
figure(1)
axis([-1000,1000,-1000,1000,-1000,1000])
axis off
%% Compute Inverse Kinematics
homeConfig = Kuka.homeConfiguration;
tform = getTransform(Kuka,homeConfig,'base','end_effector');
ik = inverseKinematics('RigidBodyTree',Kuka);
weights = [0.25 0.25 0.25 1 1 1];
initialguess = Kuka.homeConfiguration;
[config,Info] = ik('end_effector',tform,weights,initialguess)
show(Kuka,config);So you now have an inverse kinematics solver for your robot but the trajectory planning depends on how you want to do it because it can be as complicated or as simple as you want to make it, it just depends on what you are trying to achieve.
You could define a path in point space and then use your IK solver to get the joint angles at that target point and then compute the difference between your current position and your target location on the path, then step through the robot joint angles until you reach that point, putting certain constraints around them so you define the direction the axis is to move +/- and so you don't have two axis' come into contact with each other or don't rotate past the limits of the simulated motors.
With a multi DoF robot you will also have the possibility of reaching a point in various different ways so you will need to consider adding in parameters to define the way in which you want the robot to reach the point e.g. imagine reaching your hand out to grab something on your desk, you could reach the object with your elbow facing toward the ceiling but could also reach it with your elbow facing the floor.
Then to add further to that you can then dive into optimization techniques on how to optimize your path planning based on specific parameters.
Check out this video and it will help you:
I did everything but did not get anything out of it, I was constantly confused and started all over again.
I'm noob on robotics toolbox.
for example I saw all these links:
https://www.mathworks.com/help/supportpk…ndeffector.html
https://www.mathworks.com/matlabcentral/…ot-manipulators
https://www.mathworks.com/help/robotics/…idbodytree.html
https://www.mathworks.com/help/robotics/…a5-59f2ff1a4c68
https://www.mathworks.com/help/robotics/….html#d122e8296
This is that I want to do:
Did you look at the video in the link that I sent you? it has everything that you need and even has the files to download to do it
