A T-S Fuzzy Model-Based Algorithm for Blood
Glucose Control in Diabetic Patients
Zigui Kang
∗
, Juntao Pan
∗
, Weiwei Zhang
∗
, Xiaoyan Zou
†
∗
School of Electrical and Information Engineering, North Minzu University, Yinchuan, China 750021
Email: zhv2008@gmail.com
†
The First People’s Hospital of Yinchuan, Yinchuan, China
Abstract—This paper is concerned with fuzzy model-based
blood glucose control problem for diabetes patients. The glucose
metabolism model used in this paper is expressed in the Takagi-
Sugeno (T-S) form obtained by sector nonlinearity approach.
Two different rates of basal exogenous insulin are considered
to analysis and design T-S fuzzy controllers. It is shown that
conditions for the solvability of the blood glucose controller
design given here are written in the form of linear matrix
inequality (LMI) which can be efficiently solved by convex
optimization techniques. One simulation example is given to
demonstrate the validity of the proposed approach.
I. INTRODUCTION
Diabetes Mellitus is a kind of the diseases in which the
body cannot produce and respond to insulin requirement of
body adequately. The diabetes patients are certainly in high
blood glucose level which is known as hyperglycemia. When
the high level of glucose exists in our bloodstream for a long
time, it is dangerous and may cause other complications. Thus,
there have been many researchers studied to control the level
of the glucose. However, the way the blood glucose respond
to insulin requirement of body is a complex nonlinear process
which directly lead to great difficulty in controlling the level
of blood glucose.
Nonlinear control systems based on the Takagi-Sugeno (T-
S) fuzzy model have received a great attention over the last
decade [1]. The merit of such fuzzy model-based control
methodology is that it offers an effective and exact repre-
sentation of complex nonlinear systems in a compact set of
state variables. With the powerful T-S fuzzy model, a natural,
simple and systematic design control approach can be provided
to complement other nonlinear control techniques that require
special and rather involved knowledge. The paper given here
aims to establish blood glucose control conditions for diabetic
patients based on a T-S fuzzy method. The parameters involved
in controller can be computed by solving a set of LMI. An
illustrative example is given to demonstrate the effectiveness
of the proposed approach.
0
This work is supported by National Natural Science Foundation (NNSF)
of China under Grant 61463001; Natural Science Foundation of Ningxia
Hui Autonomous Region under Grant 2018AAC03106, 2018AAC03107; The
Fundamental Research Funds for the Central Universities, North Minzu
University under Grant 2018XYZDX02, 2018XYZDX03.
II. MODEL DESCRIPTION
A. Bergman minimal model
In order to quantify glucose metabolism reasonably and
accurately, a mathematical model of glucose metabolism need
to be established firstly. However, glucose metabolism is a
complex nonlinear process, subject to various perturbations,
e.g. Practice of an exercise, Meal consumptions, which makes
it really difficult to model. Several assumptions should be
done while deriving the model of glucose metabolism. The
most common are to neglect the effect of all hormones but
insulin. It is believed that the nonlinear and time continuous
aspects of the glucose metabolism are of prime importance
while designing model-based glucose control algorithm in
this paper. Therefore, it is desirable to work with Bergman
minimal model [2] by considering computation burden and
identification difficulties which is give as follows:
˙
I(t) = −nI(t) + p
5
u
1
(t)
˙
X(t) = −p
2
X(t) + p
3
(
I(t) − I
b
)
˙
G(t) = −p
1
G(t) + p
4
X(t)G(t) + p
1
G
b
+
u
2
(t)
V ol
G
(1)
where I(t) is plasma insulin concentration (
µU
mL
). X(t) is
remote insulin concentration (
µU
mL
). G(t) is plasma glucose
concentration (
mg
dL
). u
1
(t) is the exogenous insulin infusion
rate (
mU
min
). Dietary absorption or external infusion of glucose
is indicated by u
2
(t) (
mg
min
). I
b
=
p
5
n
u
1b
and G
b
are the basal
plasma insulin and glucose, respectively. The exogenous in-
sulin infusion rate to maintain I
b
is represented by u
1b
(
mU
min
).
The rate constant n represents clearance of plasma insulin.
Parameter p
5
represents the inverse of insulin distribution
space. The rates of appearance of insulin in, and disappearance
of remote insulin from, the remote insulin compartment are
governed by the parameters p
2
and p
3
, respectively. The
glucose distribution space is indicated by V ol
G
. Parameter
p
1
represents the rate at which glucose is removed from the
plasma space into the liver or into the periphery independent
of the influence of insulin. Glucose uptake under the influence
of insulin is governed by the parameter p
4
. Parameter values
for the Bergman minimal model (1) are provided in TABLE I
[3].
B. Mixed meal model
The Bergman minimal model (1) accepts meal-induced
disturbance in the form of glucose u
2
(t). A mixed meal usually
978-1-5386-4509-3/18/$31.00 ©2018 IEEE
2018 IEEE 3rd Advanced Information Technology, Electronic and Automation Control Conference(IAEAC 2018)