Advances in Intelligent Systems Research, volume 134

2nd International Conference on Control, Automation, and Artificial Intelligence (CAAI 2017)

Trajectory Planning for 6-DOF Robotic Arm Based

on Quintic Polynormial
Xiaojie Zhao1, 2, Maoli Wang1, 2, *, Ning Liu3 and Yongwei Tang1, 2
1
Shandong Computer Science Center (National Supercomputer Center in Jinan), Jinan 250014, China
2
Shandong Provincial Key Laboratory of Computer Networks, Jinan 250014, China
3
School of automation engineering, Qufu Normal University, Rizhao, 276826, China
*Corresponding author

Abstract—The paper deals with research on polynomial zi

trajectory planning for 6-DOF robotic arm. On the basis of i xi
kinematics analysis, we carried out trajectory planning and z i -1
simulation of joint space. Through the trajectory planning ai 第 i+ 1 连 杆
diagrams of MATLAB simulation on the two methods, the oi
results showed that quintic polynomial is better than cubic i 第 i+ 1 关 节
polynomial in the effect of planning. As a result, this method
第 i连 杆 di
used in six degrees of freedom industrial robot trajectory
planning is feasible. o i -1
第 i关 节 x i -1
Keyword-robotic arm; trajectory planning; 6-DOF; polynomial
FIGURE I. THE COORDINATE FRAME OF LINK
I. INTRODUCTION
TABLE I. THE D-H PARAMETERS
With the development of the economy, the application of
manipulator is more and more extensive, and its technical  i /(°)  i-1 /(°) a i-1 /(mm) di /(mm)
requirements are getting higher and higher. Using the 1 0° 0 0
mechanical arm operation can improve the quality and quantity
of the products, increase productivity and reduce cost while 2 -90° 0 280
trajectory planning directly affects the quality of the robot's 3 0° 800 0
work [1]. The trajectory refers to the displacement, velocity and 4 -90° 60 700
acceleration of the manipulator in motion. Calculating the
5 90° 0 0
expected trajectory is trajectory planning which is based on the
requirements of the task [2].The requirement of the trajectory 6 -90° 0 120
planning is to control the movement speed of the manipulator
during the movement and the movement space is always kept A. Positive Kinematics Analysis
within the allowable range of each joint movement. The The so-called positive kinematics is that the angles and
running trajectory is smooth and continuous [3]. The geometric parameters of the mechanical arm are known
mechanical arm operation experiment is carried out by using whereby the position of the robot end actuator relative to the
MATLAB Robotics Toolbox [4]. It turns out that the toolbox reference coordinate system. Positive kinematics is an
can simulate a robot. If you choose different types of joint indispensable step in the display of the movement trajectory of
interpolation functions under the required constraints, you can the mechanical arm. A homogeneous transformation matrix
generate different trajectories in the joint space for trajectory i-1Ti can be determined between the adjacent link i and the
planning. link (i-1) for the 6-DOF series robotic arm.
Using cubic polynomials and seven polynomials to
trajectory planning has some drawbacks. In order to meet the  ci si 0 ai1 
requirement that 6-DOF robotic arm should run a smooth, s c c c s d s 
continuous trajectory during the operation, we will use quintic i1
Ti  i i1 i i1 i1 i i1
(1)
polynomial. MATLAB simulation verifies the feasibility of the sisi1 cisi1 ci1 dici1 
method.  
 0 0 0 1 
II. KINEMATICS ANALYSIS
The coordinate frame of link of 6-DOF series robotic arm The transformation matrix T6 of the mechanical arm
is shown in Figure 1. The D-H parameters are shown in Table coordinate system {6} with respect to the base {0} can be
I. obtained by sequentially multiplying the homogeneous
transformation matrix i-1Ti in sequence. We can get a matrix
based on  i according to Table 1 and the formula (1).

Copyright © 2017, the Authors. Published by Atlantis Press.

This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/). 115
 Advances in Intelligent Systems Research, volume 134

B. Inverse Kinematics Analysis tasks to be accomplished by its end actuators. We can select
The inverse kinematic analysis of robots is the basis of the different joint interpolation functions to generate different
planning and control of the mechanical arm movement. The trajectories.
value of each joint angle  i is obtained by the known A. Cubic Polynomial Interpolation
end-effector, which is relative to the position and linkage We define the start point of the robotic arm joint movement
parameters of the reference coordinate system. ist0 , the start point of the joint angle is 0 , the end point istf ,
Take  1 as an example, Multiplied by the inverse of 0T1 the joint angle is  f .The description of the trajectory can be
represented by a smooth interpolation function of the start
on left side of the equation T6  0T11T2 2T3 3T4 4T5 5T6 which
0
is angle and end angle. Assume that the joint speed at both times
0 10 is known and set to zero. Then, in order to achieve a smooth
T T  1T2 2T33T4 4T55T6 ,
1 6
the inverse of 0T1 is equal to its motion of a single joint, the trajectory function (t ) satisfies
transpose. at least four constraints:

 c1 s1 0 0   (0 )   0
0 0   ( t )  
0 -1 0  0 1 
T1  T1    
f f

 s1 c1 0 0   ( 0 )  0
    ( t f )  0
0 0 0 1 (2) （7）

Then we derive The above constraint determines the functional equation of

the joint angle and velocity:
c1 s1 0 0 nx ox ax px 
0 0 1 0 ny oy ay py  1 2 3 4 5  (t )  a0  a1t  a2t 2  a3t 3
   T2 T3 T4 T5 T6  
  (t )  a1  2a2t  3a3t
2
s1 c1 0 0 nz oz az pz  (8)
   
0 0 0 1  0 0 0 1  (3) We bring the constraint (7) into the equation (8).
Expanding the formula and according to the principle of Solving the equation we can derive
equality on both sides of the matrix, we derive
 a0   0
 a1  0

 s1 p x  c1 p y  d 2  a  3 (   )
（4）  2 t2 f 0
 f

Using triangular substitutions: a  - 2 (   )

 3 t 3f
f 0
 (9)
B. Quintic Polynomial Interpolation
p x   cos  , p y   sin 
（5） In the case where the trajectory is more stringent and the
constraint condition is increased, the cubic polynomial
When   px 2  py 2 ,   a tan 2(px ,py ) . Substituting interpolation can’t satisfy the requirement, and the high order
polynomial is used for interpolation. For example, when the
(5) into (4), we can get the solution of  1 : starting point and the ending point of a certain path are
specified for the position, velocity and acceleration of their
1  atan2( py , px ) atan2(d2 , px  py d2 ) （6）
2 2 2
joints, a quintic polynomial can be used for interpolation. The
functional equation of the joint angle, velocity and acceleration
The positive and negative signs correspond to the two is as follows:
possible solutions.
Similarly, we can get the value of other x. Usually  (t )  a0  a1t  a2t 2  a3t 3  a4t 4  a5t 5
following the principle of "move the small joints more, move  
  (t )  a1  2a2t  3a3t  4a4t  5a5t
2 3 4
the large joints less" to select the most satisfying set of  (t )  2a  6a t  12a t 2  20a t 3
solutions to meet the requirements of the robotic arm.  2 3 4 5
(10)
III. JOINT SPACE TRAJECTORY PLANNING
The task of the robotic arm trajectory planning is to design
the motion function of each joint of the arm based on certain

116
 Advances in Intelligent Systems Research, volume 134

Assuming six conditions are known, i.e.,  (t 0 )   0 :  (t f )   f curves obtained by cubic polynomials, the joint acceleration
(t 0 )  (t 0 )  (t f )  (t f )  0 curves are not straight but curved, and have better effect.
.We bring these known conditions
into the form (10) and get the coefficients :

 a0   0
 a1  0

 a2  0
 10( f   0 ) (a) displacement (b) velocity (c) acceleration
 a3 
 t 3f
TABLE III. CUBIC POLYNOMIAL JOINT CURVES
  15( f   0 )
a 4 
 t 4f
 6( f   0 )
 a5 
 t 5f
(11)
IV. SIMULATION RESULTS
Set starting point p0 = [970, 0, 1140] and the target point
p1 = [611.069, 386.390, -19.873], in the working space of the
known robotic arm. The joint angle values of the joints at two
points are obtained by the ikine function in MATLAB
Robotics Toolbox as shown in Table II. Using the link function
(a) joint 1 (b) joint 2 (c) joint 3
to construct the FS20N 3D model, its position and orientation
at different points are shown in Figure 2.

TABLE II. THE JOINT ANGLE VALUES OF THE JOINTS AT TWO

POINTS

1 2 3 4 5 6
Start point 0 0 0 0 0 0
Target point 60 80 100 70 50 120

(d) joint 4 (e) joint 5 (f) joint 6

TABLE IV. QUINTIC POLYNOMIAL 1 ~ 6 JOINT TRAJECTORY CURVE

V. CONCLUSIONS
(a) Start point (b) Target point In this paper, we carry out the kinematics analysis of
FIGURE II. THE POSITION OF THE JOINTS AT TWO POINTS
6-DOF robotic arm, and the application of cubic polynomial
and quintic polynomials in 6-DOF robotic arm trajectory
The motion planning of the manipulator from the start planning is emphasized. Finally, we obtained the
point to the target point is realized by the cubic polynomial corresponding curve by MATLAB simulation. After
interpolation and the quintic polynomial polynomial comparing and analyzing, the results show that the quintic
interpolation method respectively. The angle, velocity, polynomials method has more obvious effect on the trajectory
acceleration curve is also obtained. The start point and target planning of the 6-DOF robotic arm.
point of both approaches are zero, and the acceleration of the We obtained smooth and continuous trajectory planning
start point and target points is zero too. curve through quintic polynomials, indicating that this method
Figure 3 shows the angular displacement, angular velocity is feasible in the application of trajectory planning of similar
and angular acceleration curve of each joint obtained by cubic 6-DOF robotic arm. This method can make the robotic arm to
polynomials(The joints in the figure from top to bottom are: run smoothly, reduce the occurrence of jitter and vibration,
joint 6, joint 3, joint 2, joint 4, joint 1, joint 5).It can be seen improve the stability of system, prolong the life of robotic arm.
from the figure that the angular displacement function, angular It can provides a reference for improving the smooth trajectory
velocity function and angular acceleration function of each planning of robotic arm.
joint are continuous, the movement of the robot in this
ACKNOWLEDGEMENT
movement is smooth. In Fig. 4, the joint curves obtained by the
quintic polynomial are smoother and smoother than the joint This research was financially supported by National
Science and Technology Major Project (2014ZX04015011);

117
 Advances in Intelligent Systems Research, volume 134

Shandong Province Major Special Support Funds

(2015ZDZX10002); Shandong Province Major Special
Support Funds (2015ZDXX0101G05).
REFERENCES
[1] YuYang Fu,6R.Research on Structural Design and Motion Trajectory
Planning of Industrial Robot [D].Master Thesis of Wuyi
University.2013.
[2] ZiXing Cai, Robotics [M]. Beijing: Tsinghua University Press, 2009:
264.
[3] LinShi Gao,XiaoLin Liu And YuanXian Ou,Robotic arm smoothing
programming algorithm based on quintic polynomials[J],Manufacturing
Automation, 2013, (11): 16-17, 24.
[4] Saeed B.Niku. Introduction to Robotics [M]. BEIJING: Publishing
House of Electronics Industry, 2013.
[5] Bin Xie, ZiXing Cai,Simulation experiment teaching of Robotics Based
on MATLAB Robotics Toolbox [J],Computer Education,2010

118