Verification and Application of Discrete-Time ZD
with Cube Characteristics for Solving the System of
Time-Varying Nonlinear Equations
Feng Xu, Zexin Li, and Dongsheng Guo
College of Information Science and Engineering, Huaqiao University
Xiamen 361021, China
gdongsh@hqu.edu.cn, xufeng1101 19@163.com
Abstract—In this study, to solve the system of time-varying
nonlinear equations, a discrete-time Zhang dynamics (ZD) with
cube error pattern is presented and investigated. Then, theoretical
results are given to indicate the cube characteristics of the
presented discrete-time ZD; i.e., decreasing the value of the
sampling gap by 10 times leads to the reduction of the steady-
state residual error by 1000 times. Numerical results via a
specific example are illustrated to demonstrate the efficacy of
the presented discrete-time ZD. Finally, this discrete-time ZD
is applied to a six-link planar robot manipulator by solving
the system of time-varying nonlinear kinematic equ ation s. T he
related simulation results further show the application prospect
of the presented discrete-time ZD.
Keywords–discrete-time Zhang dynamics, cube character-
istics, theoretical results, the system of time-varying nonlin-
ear equations, redundant robot manipulators.
I. INTRODUCTION
The system of nonlinear equations is involved in various
scientific and industrial fields [1]–[3]. How to solve the
system compr ising nonlin ear equ ations effectively has now
been viewed a sign ifica nt issue. Considerable efforts have been
given to find the solution to the system of nonlinear equations
[1]–[8]. For example, a Newton-based approach was investi-
gated by Ramos an d Monteiro [4]. In [5], a general class of
optimal order multi-point methods was studied by Sharma et al
to solve nonlinea r equations. Besides, focusing on the system
of nonline ar equa tions, Xiao and Yin presented the solution
that achieves higher order of convergence [6]. In compliance
with nu merical algorithms, the neurodynamic approaches were
also studied to solve the system of nonlinear equations [2], [3],
[8]. Among the neurodynamic approaches, the gradient-based
dynamics (GD), which processes the exponential convergence,
is a classic solution for nonlinear equations [2], [3].
This work is supported by the National Natural Science Foundation
of China with number 61603143, by the Promotion Program for Young
and Middle-aged Teacher in Science and Technology Research of Huaqiao
University with number ZQN-YX402, by the Scientific Research Funds of
Huaqiao University with number 15BS410, by the Natural Science Foundation
of Fujian Province with number 2016J01307, and also by the Subsidized
Project for Cultivating Postgraduates’ Innovative Ability in Scientific Research
of Huaqiao University with number 1611301044.
Copyright © 2018 IEEE. Personal use of this material is permitted.
However, permission to use this material for any other purposes must be
obtained from the IEEE by sending a request to pubs-permissions@ieee.org.
In recent years, the engineering systems in many in dustries
always change with time t. Thus, the cor responding system of
nonlinear equations should be time-varying. In this situation,
the widely-studied system of nonlin ear equations (being time-
invariant ones) is transformed into the following time-varying
system consisting of nonlinear equations [2], [3 ], [9 ]–[11]:
f(x(t), t) = 0 ∈ R
m
, ∀t ∈ [0, +∞), (1)
where f (·) : R
n
→ R
m
is a differentiable mapping and x(t) ∈
R
n
is the vector that needs to be determined.
Since 2008, Zhang et al have formally presented a typical
neurodynamic approach [10]–[13], namely Zhang dynamics
(ZD), which is considered as a superior method for solving
time-varying problems. Such a Z D differs from the GD in
the light of the error definition and the use of time-derivative
informa tion. Especially for the solution to the time-varying
system comprising nonlinear eq uations, i.e., (1), a continuous-
time ZD was studied in [10]. Aided with Euler difference rule,
[11] further investigated the discrete-time ZD. Th is discrete-
time ZD has an O(τ
2
) error pattern in solving the system of
time-varying nonlinear equations, where τ is the sam pling gap.
That is, the steady-state residual error (SSRE) of the discrete-
time ZD is redu ced by 100 times by decreasing the value of
τ by 10 times.
In this study, differing from the previous discrete-time ZD
with O(τ
2
) error pattern, a new discrete-time ZD with O(τ
3
)
error pattern is presented and investigated to find a solution
of (1). It is then theoretically proven the cube ch aracteristics
of this discrete-time ZD; th a t is, decreasing the value of τ by
10 times results in the reduction of th e SSRE by 1000 times.
With a n illustrative example, numerical results further indicate
the efficacy of the presented discre te -time ZD.
Considering the increasingly importa nt role of redundant
robot manipulators in numer ous fields [13]–[19 ], the appli-
cation of the presented discrete-time Z D is provided as well
in this study. Specifically, by solving the time -varying system
consisting o f nonlinear kinematic equations [18], [19], such
a discrete-time ZD is applied to redundant ro bot manipu-
lators. On the basis of a six-link planar robot manipulator,
simulation results indicate the application prospect of the
presented discrete-time Z D . This robot application can provide