Robotics and Autonomous Systems 105 (2018) 138–145
Contents lists available at ScienceDirect
Robotics and Autonomous Systems
journal homepage: www.elsevier.com/locate/robot
General frame for arbitrary 3R subproblems based on the POE model
Haixia Wang
a,b
, Xiao Lu
a,b
, Chunyang Sheng
b
, Zhiguo Zhang
a,b
, Wei Cui
b
, Yuxia Li
a,b,
*
a
Key Laboratory for Robot and Intelligent Technology of Shandong Province, Shandong University of Science and Technology, Qingdao, 266590, China
b
College of Electrical Engineering and Automation, Shandong University of Science and Technology, Qingdao, 266590, China
a r t i c l e i n f o
Article history:
Received 26 October 2017
Received in revised form 5 March 2018
Accepted 11 April 2018
Available online 24 April 2018
Keywords:
Inverse kinematics
Product of exponentials (POE)
Screw theory
Rodrigues’ rotation formula
a b s t r a c t
Inverse kinematic (IK) problems based on the product of exponentials (POE) model are often transformed
to a series of Paden–Kahan subproblems, which are based on geometric methods with geometric
constraints. The focus of subproblem is to solve the 3rd order subproblems, among which the 3R-type
subproblem is the most prevalent. In this paper, a general frame for the inverse solution of arbitrary 3R
types, without geometric constraints, is presented. It can be applied in different 3R types of practical cases,
including those with parallel, intersecting, and skewed relationships. This paper mainly focuses on: (1)
developing a real algebraic geometric (RAG) method based on the properties of the screw theory and the
Rodrigues’ rotation formula; (2) obtaining the closed-form solutions for arbitrary 3R subproblems, and
ensuring the accuracies of these solutions; (3) expanding the Paden–Kahan subproblems and meeting the
demands of online real-time applications; and finally, (4) verifying the effectiveness of the RAG method,
through comparisons with the geometric method, using simulations and real experiments. The proposed
frame can be widely applied in series, reconfigurable, and other types of robots.
© 2018 Published by Elsevier B.V.
1. Introduction
The most important step to make a robot follow a certain tra-
jectory or complete a grasping task is to obtain the joint angles by
solving the inverse kinematic problem [1]. This technique has been
widely used in various fields such as biomechanics [2–4], medical
robotics [5,6], bionic robotics [7,8], reconfigurable robotics [9–11],
redundant manipulator [12] and so on.
With respect to algorithms, the inverse kinematic problems
of robots are generally sorted into those with closed-form so-
lutions and the ones with numerical solutions, the closed-
form solutions being divided into geometric and algebraic solu-
tions [13]. There are many numerical methods, for example, the
iterative method [14], genetic algorithm [15,16], neural-network
method [17], calculus resolving [18], and so on. Although these
methods are considered viable solutions, and can obtain the in-
verse solution without robotic structural limits, they are time
consuming, and their completeness, convergence, and robustness
cannot be guaranteed. Therefore, we are more inclined toward
obtaining closed-form solutions of robots because of their accura-
cies and real-time capabilities. However, the conventional closed-
form inverse kinematic methods based on the Denavit–Hartenberg
(D–H) model need extensive manual derivation efforts and spe-
cific analyses for each case [19]. By contrast, the Paden–Kahan
*
Corresponding author.
E-mail address: yuxiali2004@sdust.edu.cn (Y. Li).
subproblems method based on the product of exponentials (POE)
model just needs to choose a few appropriate subproblems to
solve the given problem. Additionally, it can be applied in various
studies of robotics since of its geometric meaning and numerical
stability [20–24]. For example, Cheng [20] improved the existing
subproblem method and applied it to the Qianjiang-I robot. Vin-
cent [21] proposed a geometry-based algorithm inspired by the
Paden–Kahan subproblems, and applied it to the path planning for
steerable needles. Rafael [22] used the Paden–Kahan subproblems
to obtain an approximate solution for a nonclassical robot structure
– a human robot with a nonspherical hip – and refined the results
using the Levenberg–Marquardt algorithm. Dong [23] applied the
Paden–Kahan subproblems to the inverse kinematics of fingers.
Sariyildiz and Temeltas [24] presented a new method based on
screw theory and quaternion algebra and obtained a compact
closed form in the inverse kinematic solution.
Generally, Paden–Kahan subproblems fall into three or-
ders [25]: 1st order subproblems, 2nd order subproblems, and
3rd order subproblems. Every order of subproblem is divided into
several types; the 3rd order-subproblem, for example, includes the
RRT_TRR, RTR, TTR_RTT, TRT, TTT, and RRR types (where R denotes
rotation and T denotes translation). The 3R (short for RRR) type is
a highly complex and common case, but the existing methods for
the 3R type are limited by the geometrical relationships between
the adjacent axes in different surfaces, such as the parallel [20,25],
intersecting [26,27], and vertical [28] relationships. In practice, it is
very difficult to keep these relationships accurate owing to errors
caused by production, installation, etc., therefore, it is difficult to
https://doi.org/10.1016/j.robot.2018.04.002
0921-8890/© 2018 Published by Elsevier B.V.