Journal of Systems Engineering and Electronics
Vol. 24, No. 5, October 2013, pp.768–774
Autotuning algorithm of particle swarm PID parameter
based on D-Tent chaotic model
Min Zhu
1
, Chunling Yang
1,*
, and Weiliang Li
2
1. School of Electrical Engineering and Automation, Harbin Institute of Technology, Harbin 150001, China;
2. China Faw Group Corporation R&D Center, Changchun 130000, China
Abstract:
An improved particle swarm algorithm based on the
D-Tent chaotic model is put forward aiming at the standard par-
ticle swar m algorithm in this paper. Convergence rate of the late
of proposed algorithm is improved by revising the inertia weight
of global optimal particles and the introduction of D-Tent chaotic
sequence. Through the test of typical function and the autotuning
test of proportional-integral-derivative (PID) parameter, we finally
make a simulation to the servo control system of a permanent
magnet synchronous motor (PMSM) under double-loop control of
rotating speed and current by utilizing the chaotic particle swarm
algorithm. Studies show that the proposed algorithm can reduce
the iterative times and improve convergence rate under the condi-
tion that the global optimal solution can be got.
Keyw ords: D-Tent, particle swarm, proportional-integral-
derivative (PID) parameter optimization.
DOI: 10.1109/JSEE.2013.000XX
1. Introduction
In the field of automatic control, as one of classical con-
trol algorithms, proportional-integral-derivative (PID) con-
trollers still have enormous vitality on engineering ap pli-
cation for their simple structure, good robustness and high
reliability [1]. Generally, the design of PID controllers
mainly refers to selections of proportional coefficient K
p
,
integral coefficient K
i
and differential c oefficient K
d
.And
the performance of PID controller quite depends on the se-
lection of its parameters. In the practical application, since
nonlinear factors and uncertainties of the system can result
in the change of the system model, the original tuning pa-
rameter does not follow the change, which causes control
effect not good enough [2–4]. Although many researches
were aiming at autotuning of the PID parameter, and
Manuscript received December 14, 2011.
*Corresponding author.
This work was supported by the National Natural Science Foundation
of China (61301011) Fundamental Research Funds for the Central Uni-
versities (HIT.NSRIF.2012010) and Heilongjiang Postdoctoral Financial
Assistance (LBH-Z11157).
many parameter autotuning methods have been put for-
ward, there still more or less insufficiency exists as auto-
tuning being made by these methods [5–12].
A kind of new swarm intelligence algorithm– particle
swarm optimization (PSO) was proposed in [13]. As soon
as this algorithm was put forward, it garnered much atten-
tion of scholars in relevant fields by its simple algor ithm
structure and strong optimization capacity.
In this paper, firstly a further research is made aiming at
standard particle swarm optimization, and a kind of chaotic
article swarm optimization is put forward on the base of
it. Then the test of the typical function and the autotuning
test of the PID parameter are made for this algorithm. Fi-
nally, a simulation is made to the servo control system of
a permanent magnet synchronous motor (PMSM) under
double-loop control of rotating speed and current by uti-
lizing the chaotic particle swarm algorithm.
2. Standard PSO
Some scholars found that there are three internal rules for
the complex swarm behavior of b irds by observing the ac-
tivities of birds: far away from the nearest neighbor, ap-
proach to the target, approach to the center of swarm [14–
16]. Suppose that a swarm is searching food randomly in
the certain scope and only one block of food in the scope,
they only know how far is from the food, except the spe-
cific position of the food. By continuously searching the
surrounding scope of the bird which is the nearest to the
food, the swarm can find the food quickly. This is the pro-
totype of the PSO algorithm [17,18].
Suppose that there exist m particles in N dimen-
sional searching space, th e position for particle i (i =
1, 2,...,m) is x
i
=(x
i1
,x
i2
,...,x
iN
), and the present
flying speed of this particle is v
i
=(v
i1
,v
i2
,...,v
iN
).The
optimal solution found during flying process of this parti-
cle is called individual optimal solution, which is marked
as p
Best
, and the corresponding optimal position is marked
as p
i
=(p
i1
,p
i2
,...,p
iN
). The optimal solution found