Page 51 - Read Online

Tong et al. Intell Robot 2024;4:125-45 I http://dx.doi.org/10.20517/ir.2024.08 Page 129

Zs
X2 Z2
J2
Z0
Y0 Ys
X0
d3 Xs J3 Z3
X3
r
X1
a3
J1 Y4
Z1
X4
Z4
J4
d5
Z5
X5
Y5 J5
Figure 2. Robot model.

Table 1. MDH parameters

s 0 0 0 30
1 90 0 0 1 − 19.14712
2 -60 0 0 2 + 70.5288
3 60 0 102.74 3 − 109.4712
4 0 270 0 4 + 90
5 -90 0 -195 5
MDH: Modified Denavit-Hartenberg.

2.1 Kinematic analysis
According to the parameter table, the coordinate transformation relationship between the joints can be ex-
pressed in the form of a rotation matrix and a position matrix; the rotation matrix is given as
(1)
5 = 0 1 0 1 2 0 2 3 0 4 0 4 5
and the position absolute matrix is defined as

 0   3 
   
   
0 +
5 = 0 1 0 1 2 0 − 2 3  + 0 1 0 1 2 0 2 3 0  

   
0
 2 3   
(2)
 0 
 
 
0 1 0 1 2 0 2 3 0 4 0 5 

 
0
 
Thederivationandtransformationoftheforwardandinversekinematicsofarobotcanbeachievedthroughthe
rotation and position matrices, realising the mutual mapping between the exoskeleton robot’s joint space and
under the Cartesian space. Through the forward and inverse kinematics resolution, it completes the execution
ofthetrainedmotionalongthepre-setdesiredtrajectory, forexample, shoulderjointadductionandabduction,
forward and backward flexion and extension, internal and external rotation movements, elbow flexion and
extension movements, and wrist turning movements. In the passive training mode, the purpose of the robot
controller is to reduce the trajectory tracking error, so that the patient learns the correct movement pattern,
usually in the form of traditional closed-loop control combined with feedforward compensation.
2.2 Identification of dynamical model parameters
Whether it is the feedforward compensation function in passive training or the interactive force-assisted con-
trol function in active training, it is necessary to establish the kinetic model of the rehabilitation robot to
calculate the moment information in the motion state, and this paper adopts the method of kinetic identifica-
tion to obtain the kinetic parameters of the robot and complete the construction of the model. Because the