Abstract—In this paper a medical robot based on a 3-RPS
parallel platform is presented for providing correct screw path
in transpedicular fixation surgery. For improving the accuracy
of the robot, a calibration approach for 3-RPS with CMM
(Coordinate Measuring Instrument) data is developed.
Sensitivity analysis of the error source had been done with
Monte-Carlo techniques. The results show that the length errors
of the
joint-links are main contribution to the output error of
the 3-RPS parallel manipulator; and the secondary factor is the
radius error of the static and the moving platform. Based on
above analysis, the calibration model of the inverse kinematics
was established and the mono-branched calibration method was
used. The presented method has high identification accuracy for
length error of the chain. Based on the identification parameters,
the results show that the method can improve the 3-RPS output
accuracy.
Keywords—parallel manipulator; accuracy analysis;
Monte-Carlo techniques; calibration
I. INTRODUCTION
In recent years, Computer Assisted Minimal Invasive
Surgery (CAMIS) becomes a research focus. In minimal
invasive spinal surgeries, such as transpedicular fixation
surgery, three-dimensional navigation equipment has been
widely used, so that surgeons can get the surgery information
based on three-dimensional image displayed on navigation
device [1, 2]. But the surgery is still manual operation by
doctors. The accuracy of operation is greatly influenced by
human factors such as tiredness and physiological tremor.
Therefore, lots of scholars are focusing on medical robotics to
improve the accuracy of spinal surgery.
Parallel manipulator with high precision, rigidity, fast
response and many other advantages are widely emerging in
high precision required fields, as spinal surgery. A typical
instance is the Spine-Assist system developed by Mazor
(Israel), which is based on Stewart platform, and used for
transpedicular fixation surgeries. In cervical disc replacement
surgery, a six Degree of Freedoms (DoFs) parallel robot is
developed by Harbin Institute of Technology (China). But for
* This research supported by the National Natural Science Foundation of
China (No.61175124 and No.61210013) and Guangdong Innovative
Research Team Program (No. 201001D0104648280).
Baoqiang Guo and Haiyang Jin are with Harbin Institutes of Technology
Shenzhen Graduate School. Guangdong Provincial Key Laboratory of
Robotics and Intelligent System, Shenzhen Institutes of Advanced
Technology, Chinese Academy of Sciences. The Chinese University of Hong
Kong. (E-mail: bq.guo@siat.ac.cn and hy.jin@siat.ac.cn).
Peng Zhang and Ying Hu*(corresponding author) are with Guangdong
Provincial Key Laboratory of Robotics and Intelligent System, Shenzhen
Institutes of Advanced Technology, Chinese Academy of Sciences. The
Chinese University of Hong Kong. (E-mail: zhangpeng@siat.ac.cn and
ying.hu@siat.ac.cn
).
Hong Zhang*(corresponding author) is with Peking university Shenzhen
hospital.(E-mail: zhanghong79@yahoo.com)
Jianwei Zhang is with University of Hamburg, Germany (E-mail:
zhang@informatik.uni-hamburg.de ).
multi-DoFs parallel robots, there are some disadvantages,
such as small workspace, complex control system, difficulties
for precision compensation, etc. By contract, for less DoFs
parallel manipulators, in addition to the advantages of parallel
manipulators, they have following other advantages: simple
structure, larger workspace, easier to control, more flexible
DoFs configuration, etc. Gradually, it’s attracting more
attention of scholars. For example, Mazor developed Smart
Assist system which is a kind of four DoFs parallel
manipulators.
For less DoFs parallel manipulator with precision
requirements, manufacturing and assembling tolerances are
significant issues. Error exists because differences between
designed parameters and actual parameters, and greatly
reduce the accuracy of the parallel robot and restrict further
development of parallel robot in clinical applications. So it’s
necessary to develop a calibration method to improve the
accuracy of the parallel manipulator. Manipulator calibration
inculdes four steps: kinematic modeling, measurment of pose,
identification of kinematic parameter, error compensation [3].
Wang et al. established the branch chain kinematics equations
by D-H method of Stewart platform, and introduce error of
U-joint and the
joint-link trains [4]. There is also a direct
differential method of the input output equations about the
modeling of a parallel manipulator [5]. Horie et al. give a
classification of tolerances in spatial four-bar linkage
manipulators and show the position contacts between pairing
elements [6]. Tools used for calibration also affect the
accuracy of the calibration. Amirat and Mau-rine used a
vision system to calibrate a parallel robot [7, 8]. Meng et al.
proposed a method using a laser tracker for calibration [9].
This paper developed a medical robot based on 3-RPS
parallel platform for providing correct screw path in
transpedicular fixation surgery. For the calibration of the
manipulator the paper focuses on the work as follows: firstly,
a detailed analysis about various sources of error of 3-RPS
parallel robot is performed, and the parallel model including
those error sources is built with the D-H method. Then, a
quantitative sensitivity analysis of those error sources is
carried out by Monte-Carlo techniques [10]. According to the
sensitivity analysis, a sensitive source of error is determined
for establishing calibration model based on inverse
kinematics [11, 12]. Finally, CMM is used for calibration of
the 3-RPS parallel robot, and the experimental results show
the improvement via the calibration.
The overall structure of this article is organized as follow.
Section II presents the analysis about error sources and error
sensitivity of the 3-RPS parallel robot. The calibration model
of the robot, which contains a variety of sensitive error
sources, and parameter identification algorithm are described
in section III. Section IV presents data measurement and
acquisition of homogeneous transformation matrices.
Accuracy Analysis and Calibration of a Parallel Guidance Device for
Minimal Invasive Spinal Surgery
Baoqiang Guo, Haiyang Jin, Peng Zhang, Jianwei Zhang, Ying Hu*, Hong Zhang*
978-1-4799-2744-9/13/$31.00 ©2013 IEEE
Proceeding of the IEEE
International Conference on Robotics and Biomimetics (ROBIO)
Shenzhen, China, December 2013