Deterministic Stabilization of a Dynamical System using a Computational Approach

International Journal of Advanced Engineering, Management and Science (IJAEMS) [Vol-4, Issue-1, Jan- 2018]
https://dx.doi.org/10.22161/ijaems.4.1.6 ISSN: 2454-1311
www.ijaems.com Page | 24
Deterministic Stabilization of a Dynamical System
using a Computational Approach
Isobeye George1
, Jeremiah U. Atsu2
, Enu-Obari N. Ekaka-a3
1
Department of Mathematics/Statistics, Ignatius Ajuru University of Education, PMB 5047, Port Harcourt, Nigeria,
2
Department of Mathematics/Statistics, Cross River University of Technology, Calabar, Nigeria,
3
Department of Mathematics, Rivers State University, Port Harcourt, Nigeria.
Abstract— The qualitative behavior of a multi-parameter
dynamical system has been investigated. It is shown that
changes in the initial data of a dynamical system will affect
the stabilization of the steady-state solution which is
originally unstable. It is further shown that the stabilization
of a five-dimensional dynamical system can be used as an
alternative method of verifying qualitatively the concept of
the stability of a unique positive steady-state solution. These
novel contributions have not been seen elsewhere; these are
presented and discussed in this paper.
Keywords— Deterministic, stabilization, dynamical
system, steady-state solution, changing initial data.
I. INTRODUCTION
Agarwal and Devi (2011) studied in detail the mathematical
analysis of a resource-dependent competition model using
the method of local stability analysis. Other relevant
mathematical approaches to the concept of stability analysis
have been done [Rescigno (1977);Hallam, Clark and Jordan
(1983);Hallam, Clark and Lassiter (1983);Hallam and Luna
(1984);Freedman and So (1985); Lancaster and
Tismenetssky (1985);De Luna and Hallam (1987); Zhien
and Hallam (1987);Freeman and Shukla (1991);Huaping
and Zhien (1991);Garcia-Montiel and Scatena (1994);
Chattopadhyay (1996);Hsu and Waltman (1998); Dubey
and Hussain (2000);Hsu, Li and Waltman (2000);Thieme
(2000); Shukla, Agarwal, Dubey and Sinha (2001); Ekaka-a
(2009); Shukla, Sharma, Dubey and Sinha (2009); Yan and
Ekaka-a (2011);Dhar, Chaudhary and Sahu (2013);
Akpodee and Ekaka-a (2015)]. The method of thispresent
study uses the technique of a numerical simulation to
quantify the qualitative characteristics of a complex
dynamical system with changing initial data.
II. MATHEMATICAL FORMULATION
We have considered the following continuous multi-
parameter system of nonlinear first order ordinary
differential equations indexed by the appropriate initial
conditions according to Agarwal and Devi (2011):
𝑑𝑥1
𝑑𝑡
= 𝑎1 𝑥1 − 𝑎2 𝑥1
2
− 𝛼𝑥1 𝑥2 + 𝛼1 𝑥1 𝑅 − 𝑘1 𝛿1 𝑥1 𝑇, 𝑥1(0) =
𝑥10 ≥ 0, (1a)
𝑑𝑥2
𝑑𝑡
= 𝑏1 𝑥2 − 𝑏2 𝑥2
2
− 𝛽𝑥1 𝑥2 + 𝛽1 𝑥2 𝑅 − 𝑘2 𝛿2 𝑥2 𝑇, 𝑥2(0) =
𝑥20 ≥ 0, (1b)
𝑑𝑅
𝑑𝑡
= 𝑐1 𝑅 − 𝑐2 𝑅2
− 𝛼1 𝑥1 𝑅 − 𝛽1 𝑥2 𝑅 − 𝑘𝛾𝑅𝑇, 𝑅(0) = 𝑅0 ≥
0, (1c)
𝑑𝑃
𝑑𝑡
= 𝜂𝑥1 + 𝜂𝑥2 − (𝜆0 + 𝜃)𝑃, 𝑃(0) = 𝑃0 ≥ 0, (1d)
𝑑𝑇
𝑑𝑡
= 𝑄 + 𝜇𝜃𝑃 − 𝛿0 𝑇 − 𝛿1 𝑥1 𝑇 − 𝛿2 𝑥2 𝑇 − 𝛾𝑅𝑇, 𝑇(0) =
𝑇0 ≥ 0, (1e)
where
𝑥1 and 𝑥2 are the densities of the first and second competing
species, respectively, R is the density of resource biomass,
P is the cumulative concentration of precursors produced by
species forming the toxicant, T is the concentration of the
same toxicant in the environment under consideration, Q is
the cumulative rate of emission of the same toxicant into the
environment from various external sources, 𝑎1 and 𝑏1 are
the intrinsic growth rates of the first and second species,
respectively, 𝑎2 and 𝑏2are intraspecific interference
coefficients,𝛼, 𝛽 are the interspecific interference
coefficients of first and second species, respectively,𝛼1 and
𝛽1 are the growth rate coefficients of first and second
species, respectively due to resource biomass. 𝑘1, 𝑘2 and 𝑘
are fractions of the assimilated amount directly affecting the
growth rates of densities of competing species and resource
biomass, 𝜂 is the growth rate coefficient of the cumulative
concentration of precursors. 𝜆0 is its depletion rate
coefficient due to natural factors whereas 𝜃 is the depletion
rate coefficient caused by its transformation into the same
toxicant of concentration 𝑇. 𝜇 is the rate of the formation of
the toxicant from precursors of competing species. 𝛿1, 𝛿2
and 𝛾 are the rates of depletion of toxicant in the
environment due to uptake of toxicant by species and their
resource biomass, respectively.
It is assumed that the resource biomass grows logistically
with the supply rate of the external resource input to the
system by constant 𝑐1 and its density reduces due to certain

degradation factors present in the environment at a rate 𝑐2.It
is further assumed that the toxicant in the environment is
washed out or broken down with rate 𝛿0.
Research Question
For the purpose of this study, we have considered the
following vital research question:How does a given
dynamical multi-parameter system of continuous nonlinear
first-order ordinary differential equations respond to a
qualitative characteristic, that is, assuming a point
(𝑥1𝑒, 𝑥2𝑒, 𝑅 𝑒, 𝑃𝑒, 𝑇𝑒) is an arbitrary steady-state solution, as
the independent variable 𝑡 approaches infinity (𝑡 → ∞), do
𝑥1 → 𝑥1𝑒, 𝑥2 → 𝑥2𝑒, 𝑅 → 𝑅 𝑒, 𝑃 → 𝑃𝑒, 𝑇 → 𝑇𝑒 under some
simplifying initial conditions? This mathematical idea is a
necessary and sufficient condition that quantifies the
concept of the stabilization of the steady-state solution
(𝑥1𝑒, 𝑥2𝑒, 𝑅 𝑒, 𝑃𝑒, 𝑇𝑒) (Yan and Ekaka-a, 2011). In other
words, for a complex system of nonlinear first-order
ordinary differential equations whose interacting functions
are continuous and partially differentiable, what is the likely
qualitative characteristic of such a system? The focus of this
chapter will tackle the following proposed problem that is
clearly defined next.
Research Hypothesis
It is a well-established ecological fact that the initial
ecological data, which mathematicians called initial
conditions, are not static characteristic values of a
dynamical system. The corresponding core research
question is, when the initial data change, how does the
dynamical system respond to this change over a longer
period of time? This hypothesis if successfully tested and
proved in this research, has the potential to provide an
insight in the further study of ecosystem stability and
ecosystem planning.
Method of Analysis
A well-defined MATLAB ODE45 function has been used
to construct tables to determine the effect of changing
values of initial data on the stability of the dynamical
system for large values of the independent variable 𝑡.
Following Agarwal and Devi (2011), the values of
parameter values which are used in the numerical
simulations for system (1) are:
𝑎1 = 5, 𝑎2 = 0.22, 𝛼 = 0.007, 𝛼1 = 0.2, 𝑘1 = 0.1,
𝛿1 = 0.05, 𝑏1 = 3, 𝑏2 = 0.26, 𝛽 = 0.008, 𝛽1 =
0.04, 𝑘2 = 0.2, 𝛿2 = 0.04, 𝜂 = 0.5, 𝜆0 = 0.01,
𝜃 = 3, 𝜇 = 0.2, 𝛿0 = 7, 𝛾 = 0.3, 𝑐1 = 10,
𝑐2 = 0.3, 𝑘 = 0.1, 𝑄 = 30.
III. RESULTS AND DISCUSSIONS
Some twenty (20) numerical simulations are observed to
determine the effect of changing values of initial data on the
stability of the dynamical system for 𝑡 = 3650 days as
shown in Table 1 below:
Table.1: Numerical simulation of a dynamical system for changing initial data at 𝑡 = 3650 days, using a MATLAB ODE45
numerical scheme.
Example Initial Data (ID) Independent
Variable t days
x1e x2e Re Pe Te
1 1 3650 25.4443 15.2308 30.7270 6.7195 2.1454
2 2 3650 25.3091 15.2308 30.6851 6.7195 2.1086
3 3 3650 25.3551 15.2308 30.6872 6.7195 2.1054
4 4 3650 25.3783 15.2308 30.6901 6.7195 2.1042
5 5 3650 25.3923 15.2308 30.6931 6.7195 2.1041
6 6 3650 25.4018 15.2308 30.6958 6.7195 2.1046
7 7 3650 25.4085 15.2308 30.6983 6.7195 2.1055
8 8 3650 25.4129 15.2308 30.6979 6.7195 2.1066
9 9 3650 25.4169 15.2308 30.6997 6.7195 2.1080
10 10 3650 25.4202 15.2308 30.7014 6.7195 2.1094
11 11 3650 25.4229 15.2308 30.7030 6.7195 2.1108
12 12 3650 25.4252 15.2308 30.7045 6.7195 2.1123
13 13 3650 25.4272 15.2308 30.7058 6.7195 2.1138
14 14 3650 25.4288 15.2308 30.7071 6.7195 2.1153
15 15 3650 25.4303 15.2308 30.7083 6.7195 2.1168
16 16 3650 25.4316 15.2308 30.7094 6.7195 2.1183

Example Initial Data (ID) Independent
Variable t days
x1e x2e Re Pe Te
17 17 3650 25.4328 15.2308 30.7105 6.7195 2.1197
18 18 3650 25.4338 15.2308 30.7115 6.7195 2.1211
19 19 3650 25.4347 15.2308 30.7125 6.7195 2.1225
20 20 3650 25.4356 15.2308 30.7134 6.7195 2.1239
where
ID 1 = (2, 0.01, 0.01, 0.1, 0.1), ID 2 = (0.10, 0.01, 0.01, 0.1,
0.1),
ID 3 = (0.15, 0.01, 0.01, 0.1, 0.1), ID 4 = (0.20, 0.01, 0.01,
0.1, 0.1),
ID 5 = (0.25, 0.01, 0.01, 0.1, 0.1), ID 6 = (0.30, 0.01, 0.01,
0.1, 0.1),
ID 7 = (0.35, 0.01, 0.01, 0.1, 0.1), ID 8 = (0.40, 0.01, 0.01,
0.1, 0.1),
ID 9 = (0.45, 0.01, 0.01, 0.1, 0.1), ID 10 = (0.50, 0.01, 0.01,
0.1, 0.1),
ID 11 = (0.55, 0.01, 0.01, 0.1, 0.1), ID 12 = (0.60, 0.01,
0.01, 0.1, 0.1),
ID 13 = (0.65, 0.01, 0.01, 0.1, 0.1), ID 14 = (0.70, 0.01,
0.01, 0.1, 0.1),
ID 15 = (0.75, 0.01, 0.01, 0.1, 0.1), ID 16 = (0.80, 0.01,
0.01, 0.1, 0.1),
ID 17 = (0.85, 0.01, 0.01, 0.1, 0.1), ID 18 = (0.90, 0.01,
0.01, 0.1, 0.1),
ID 19 = (0.95, 0.01, 0.01, 0.1, 0.1), ID 20 = (1, 0.01, 0.01,
0.1, 0.1).
Considering Table 1, we deduced mathematically that, as
𝑡 → ∞ for the given initial conditions,𝑥1 → 𝑥1𝑒, 𝑥2 → 𝑥2𝑒,
𝑅 → 𝑅 𝑒, 𝑃 → 𝑃𝑒, 𝑇 → 𝑇𝑒. We have shown that as the initial
data are changing, the system is approaching its steady-
state. This shows that changes in the initial data of a
dynamical system will affect the stabilization of the steady-
state solution which is originally unstable.
Table.2: Test of stability of steady-state solutions for changing values of initial data, using a MATLAB ODE45 numerical
scheme.
Example Initial
Data
(ID)
Steady-state
solution (or
point)
λ1 λ2 λ3 λ4 λ5 TOS
1 1 1 -18.1705 -9.3804 -5.7543 -3.0150 -4.0180 Stable
2 2 2 -18.1498 -9.3527 -5.6958 -3.0150 -4.0180 Stable
3 3 3 -18.1527 -9.3547 -5.7159 -3.0150 -4.0185 Stable
4 4 4 -18.1547 -9.3568 -5.7260 -3.0150 -4.0186 Stable
5 5 5 -18.1563 -9.3589 -5.7321 -3.0150 -4.0187 Stable
6 6 6 -18.1576 -9.3607 -5.7362 -3.0150 -4.0187 Stable
7 7 7 -18.1587 -9.3623 -5.7391 -3.0150 -4.0186 Stable
8 8 8 -18.1588 -9.3621 -5.7411 -3.0150 -4.0187 Stable
9 9 9 -18.1596 -9.3633 -5.7428 -3.0150 -4.0187 Stable
10 10 10 -18.1604 -9.3644 -5.7442 -3.0150 -4.0187 Stable
11 11 11 -18.1610 -9.3654 -5.7453 -3.0150 -4.0186 Stable
12 12 12 -18.1616 -9.3664 -5.7463 -3.0150 -4.0186 Stable
13 13 13 -18.1622 -9.3672 -5.7472 -3.0150 -4.0186 Stable
14 14 14 -18.1627 -9.3680 -5.7478 -3.0150 -4.0185 Stable
15 15 15 -18.1632 -9.3688 -5.7485 -3.0150 -4.0185 Stable
16 16 16 -18.1637 -9.3695 -5.7490 -3.0150 -4.0185 Stable
17 17 17 -18.1641 -9.3701 -5.7495 -3.0150 -4.0185 Stable
18 18 18 -18.1645 -9.3708 -5.7500 -3.0150 -4.0184 Stable
19 19 19 -18.1649 -9.3714 -5.7505 -3.0150 -4.0184 Stable
20 20 20 -18.1653 -9.3720 -5.7507 -3.0150 -4.0184 Stable

where
Point 1 = (25.4443, 15.2308, 30.7270, 6.7195, 2.1454),
Point 2 = (25.3091, 15.2308, 30.6851, 6.7195, 2.1086),
Point 3 = (25.3551, 15.2308, 30.6872, 6.7195, 2.1054),
Point 4 = (25.3783, 15.2308, 30.6901, 6.7195, 2.1042),
Point 5 = (25.3923, 15.2308, 30.6931, 6.7195, 2.1041),
Point 6 = (25.4018, 15.2308, 30.6958, 6.7195, 2.1046),
Point 7 = (25.4085, 15.2308, 30.6983, 6.7195, 2.1055),
Point 8 = (25.4129, 15.2308, 30.6979, 6.7195, 2.1066),
Point 9 = (25.4169, 15.2308, 30.6997, 6.7195, 2.1080),
Point 10 = (25.4202, 15.2308, 30.7014, 6.7195, 2.1094),
Point 11 = (25.4229, 15.2308, 30.7030, 6.7195, 2.1101),
Point 12 = (25.4252, 15.2308, 30.7045, 6.7195, 2.1123),
Point 13 = (25.4272, 15.2308, 30.7058, 6.7195, 2.1138),
Point 14 = (25.4288, 15.2308, 30.7071, 6.7195, 2.1153),
Point 15 = (25.4303, 15.2308, 30.7083, 6.7195, 2.1168),
Point 16 = (25.4316, 15.2308, 30.7094, 6.7195, 2.1183),
Point 17 = (25.4328, 15.2308, 30.7105, 6.7195, 2.1197),
Point 18 = (25.4338, 15.2308, 30.7115, 6.7195, 2.1211),
Point 19 = (25.4347, 15.2308, 30.7125, 6.7195, 2.1225),
Point 20 = (25.4356, 15.2308, 30.7134, 6.7195, 2.1239).
What do we learn from Table 2? On the basis of this
sophisticated computational approach which we have not
seen elsewhere, we hereby infer that the stabilization of a
five-dimensional dynamical system can be used as an
alternative method of verifying qualitatively the concept of
the stability of a unique positive steady-state solution which
could have been a daunting task to resolve analytically.
However, this key contribution is only valid as long as the
intrinsic growth rate 𝑎1 is bigger than the intra-competition
coefficient 𝑎2 of the first competing species; the intrinsic
growth rate 𝑏1 is bigger than the intra-competition
coefficient 𝑏2 of the second competing species and the
intrinsic growth rate of the resource biomass 𝑐1 is bigger
than the intra-competition coefficient 𝑐2 of the resource
biomass. In the event that the intra-competition coefficients
of these three populations outweigh their corresponding
intrinsic growth rates, will the specified steady-state
solutions still be stable? Without loss of generality, it is
interesting to observe that each of the twenty (20) stable
steady-state solutions is also qualitatively well-defined
within the choice of the model dynamics in the absence of
proper model parameter estimation. The idea is consistent
with the earlier discovery of Ekaka-a (2009).
IV. CONCLUSION AND RECOMMENDATION
We have shown in this research that stabilization is an
alternative way of testing for stability. Therefore, the
application of a computational approach in the
determination of the stability characteristic using the
concept of stabilization is one of the contributions of this
work that can be used to move the frontier of knowledge in
the field of numerical mathematics with respect to stability
of a dynamical system.We recommend a further
investigation of the effect of fixed initial data for changing
values of the independent variable.
REFERENCES
[1] Agarwal, M. and Devi, S. (2011). A resource-
dependent competition model: Effects of
toxicant emitted from external sources as well as
formed by precursors of competing species. Nonlinear
Analysis: Real World Application, 12, 751−766.
[2] Akpodee, R. E. and Ekaka-a, E. N. (2015).
Deterministic stability analysis using a numerical
simulation approach, Book of Proceedings – Academic
Conference Publications and Research International
on Sub-Sahara African Potentials in the new
Millennium, 3(1).
[3] Chattopadhyay, J. (1996). Effects of toxic substances
on a two-species competitive system, Ecological
Model, 84, 287 – 289.
[4] Dhar, J., Chaudhary, M. and Sahu, G. P. (2013).
Mathematical model of depletion of forestry resource,
effect of synthetic-based industries. International
Journal of Biological, Life Science and Engineering,
7, 1 – 5.
[5] De Luna, J. T. and Hallam, T. G. (1987). Effects of
toxicants on populations: a qualitative approach IV.
Resource-consumer-toxicant model, Ecological
Model, 35, 249 – 273.
[6] Dubey, B. and Hussain, J. (2000). A model for
allelopathic effect on two competing species,
Ecological Model, 129, 195 – 207.
[7] Ekaka-a, E. N. (2009). Computational and
mathematical modeling of plant species interactions in
a harsh climate, Ph. D Thesis, University of Liverpool
and University of Chester, United Kingdom.
[8] Freedman, H. I. and Shukla, J. B. (1991). Models for
the effect of toxicant in single species and
predator−prey systems, Journal of Mathematical
Biology, 30, 15−30.
[9] Freedman, H. I. and So, J. W. H. (1985). Global
stability and persistence of simple food chains,
Mathematical Bioscience, 76, 69 − 86.
[10]Garcia-Montiel, D. C. and Scantena, F. N. (1994). The
effects of human activity on the structure composition
of a tropical forest in Puerto Rico, Forest Ecological
Management, 63, 57 – 58.
[11]Hallam, T. G., Clark, C. E., and Jordan, G. S. (1983).
Effects of toxicants on populations: a qualitative

approach II. First order kinetics, Journal of
Mathematical Biology, 18, 25 – 37.
[12]Hallam, T. G., Clark, C. E., and Lassiter, R. R. (1983).
Effects of toxicants on populations: a qualitative
approach I. Equilibrium environmental exposure.
Ecological Model, 18, 291−304.
[13]Hallam T. G. and De Luna, J. T. (1984). Effects of
toxicants on populations: a qualitative approach III.
Environmental and food chain pathways, Journal of
Theoretical Biology, 109, 411 – 429.
[14]Hsu, S. B., Li, Y. S. and Waltman, P. (2000).
Competition in the presence of a lethal external
inhibitor, Mathematical Bioscience,, 167, 177 – 199.
[15]Hsu, S. B. and Waltman, P (1998). Competition in the
chemostat when one competitor produces a toxin,
Japan Journal of Industrial and Applied
Mathematics, 15,471–490.
[16]Huaping, L, and Zhien, M. (1991). The threshold of
survival for system of two species in a polluted
environment, Journal of Mathematical Biology, 30, 49
– 61.
[17]Lancaster, P. L. and Tismenetsky, M (1985).The
Theory of Matrices, Second Edition, Academic Press,
New York.
[18]Rescigno, A. (1977). The struggle for life – V. One
species living in a limited environment, Bulleting of
Mathematical Biology, 39, 479 – 485.
[19]Shukla, J. B., Agarwal, A., Dubey, B. and Sinha, P.
(2001). Existence and survival of two competing
species in a polluted environment: a mathematical
model, Journal of Biological Systems, 9(2), 89 – 103.
[20]Shukla, J. B., Sharma, S., Dubey, B. and Sinha, P.
(2009). Modeling the survival of a resource-dependent
population: Effects of toxicants (pollutants) emitted
from external sources as well as formed by its
precursors, Nonlinear Analysis: Real World
Application, 54−70.
[21]Thieme, H. R. (2000). Uniform persistence and
permanence for non-autonomous semi-flows in
population biology, Mathematical Bioscience,
166,173 − 201.
[22]Yan, Y. and Ekaka-a, E. N. (2011). Stabilizing a
mathematical model of population system, Journal of
the Franklin Institute, 348(10), 2744 – 2758.
[23]Zhien, M. and Hallam, T. G. (1987). Effects of
parameter fluctuations on community survival.
Mathematical Bioscience, 86, 35 – 49.

Deterministic Stabilization of a Dynamical System using a Computational Approach

More Related Content

What's hot (15)

Similar to Deterministic Stabilization of a Dynamical System using a Computational Approach (20)

Recently uploaded (20)

Deterministic Stabilization of a Dynamical System using a Computational Approach