Global stability analysis of a viral infection model in a critical case

Wei Wang; Xiulan Lai; Wei Wang; Xiulan Lai

doi:10.3934/mbe.2020074

Mathematical Biosciences and Engineering

2020, Volume 17, Issue 2: 1442-1449. doi: 10.3934/mbe.2020074

Previous Article Next Article

Research article

Global stability analysis of a viral infection model in a critical case

Wei Wang ^{1
,
,},
Xiulan Lai ²

1.
College of Mathematics and Systems Science, Shandong University of Science and Technology, Qingdao 266590, China
2.
Institute for Mathematical Sciences, Renmin University of China, Beijing 100872, China

Received: 05 October 2019 Accepted: 17 November 2019 Published: 28 November 2019

Recently, it has been proved that for the diffusive viral infection model with cell-to-cell infection, the virus-free steady state E₀ is globally attractive when the basic reproduction number R₀ < 1, and the virus is uniformly persistent if R₀ > 1. However, the global stability analysis in the critical case of R₀ = 1 is not given due to a technical difficulty. For the diffusive viral infection model including a single equation with diffusion term, global stability analysis in the critical case has been performed by constructing Lyapunov functions. Unfortunately, this method is not applicable for two or more equations with diffusion terms, which was left it as an open problem. The present study is devoted to solving this open problem and shows that E₀ is globally asymptotically stable when R₀ = 1 for three equations with diffusion terms by means of Gronwall's inequality, comparison theorem and the properties of semigroup.

Keywords:

Citation: Wei Wang, Xiulan Lai. Global stability analysis of a viral infection model in a critical case[J]. Mathematical Biosciences and Engineering, 2020, 17(2): 1442-1449. doi: 10.3934/mbe.2020074

Related Papers:

[1]	Zulqurnain Sabir, Hafiz Abdul Wahab, Juan L.G. Guirao . A novel design of Gudermannian function as a neural network for the singular nonlinear delayed, prediction and pantograph differential models. Mathematical Biosciences and Engineering, 2022, 19(1): 663-687. doi: 10.3934/mbe.2022030
[2]	Zulqurnain Sabir, Muhammad Asif Zahoor Raja, Abeer S. Alnahdi, Mdi Begum Jeelani, M. A. Abdelkawy . Numerical investigations of the nonlinear smoke model using the Gudermannian neural networks. Mathematical Biosciences and Engineering, 2022, 19(1): 351-370. doi: 10.3934/mbe.2022018
[3]	Xin-You Meng, Yu-Qian Wu . Bifurcation analysis in a singular Beddington-DeAngelis predator-prey model with two delays and nonlinear predator harvesting. Mathematical Biosciences and Engineering, 2019, 16(4): 2668-2696. doi: 10.3934/mbe.2019133
[4]	Reymundo Itzá Balam, Francisco Hernandez-Lopez, Joel Trejo-Sánchez, Miguel Uh Zapata . An immersed boundary neural network for solving elliptic equations with singular forces on arbitrary domains. Mathematical Biosciences and Engineering, 2021, 18(1): 22-56. doi: 10.3934/mbe.2021002
[5]	Xinran Zhang, Yongde Zhang, Jianzhi Yang, Haiyan Du . A prostate seed implantation robot system based on human-computer interactions: Augmented reality and voice control. Mathematical Biosciences and Engineering, 2024, 21(5): 5947-5971. doi: 10.3934/mbe.2024262
[6]	Kelum Gajamannage, Erik M. Bollt . Detecting phase transitions in collective behavior using manifold's curvature. Mathematical Biosciences and Engineering, 2017, 14(2): 437-453. doi: 10.3934/mbe.2017027
[7]	Wenlong Wang, Chunrui Zhang . Bifurcation of a feed forward neural network with delay and application in image contrast enhancement. Mathematical Biosciences and Engineering, 2020, 17(1): 387-403. doi: 10.3934/mbe.2020021
[8]	Jitai Liang, Junjie Wei . Lyapunov functional for virus infection model with diffusion and state-dependent delays. Mathematical Biosciences and Engineering, 2019, 16(2): 947-966. doi: 10.3934/mbe.2019044
[9]	Xu Song, Jingyu Li . Asymptotic stability of spiky steady states for a singular chemotaxis model with signal-suppressed motility. Mathematical Biosciences and Engineering, 2022, 19(12): 13988-14028. doi: 10.3934/mbe.2022652
[10]	Yongquan Zhou, Yanbiao Niu, Qifang Luo, Ming Jiang . Teaching learning-based whale optimization algorithm for multi-layer perceptron neural network training. Mathematical Biosciences and Engineering, 2020, 17(5): 5987-6025. doi: 10.3934/mbe.2020319

Abstract

1. Introduction

The historical Lane-Emden model was introduced first time by astrophysicist Jonathan Homer Lane and Robert Emden ^[1,2] working on the thermal performance of a spherical cloud of gas and classical law linked to thermodynamics ^[3]. The singular models have several applications in broad field of applied science and engineering such as catalytic diffusion reactions along with error estimate problems ^[4], density profile of gaseous star ^[5], stellar configuration ^[6], spherical annulus ^[7], isotropic continuous media ^[8], the theory of electromagnetic ^[9] and morphogenesis ^[10]. It is always not easy to solve the system of singular equations-based models due to their complex nature and singular points. To mention a few schemes that have been applied to solve such models include Legendre wavelets spectral technique ^[11], Bernoulli collocation scheme ^[12], variational iteration technique ^[13], Haar wavelet quasilinearization method ^[14], spectral collocation scheme ^[15], differential transformation approach ^[16] and Adomian decomposition technique ^[17].

All these above cited approaches have their precise merits and imperfections, however stochastic solver has not been extensively implemented to solve multi-singular third kind of nonlinear system (MS-TKNS) using the artificial neural networks (ANNs) together with particle swarm optimization (PSO) and interior-point algorithm (IPA), i.e., ANN-PSO-IPA. The stochastic computing solvers have been widely applied to resolve numerous applications ^{[18,19,20,21,22]}. Recently, the stochastic solvers presented the solution of models for financial market forecasting ^[23], prey-predator nonlinear system ^[24], nonlinear singular functional differential model ^[25,26], SITR nonlinear system ^[27,28], singular delay differential system ^[29], nonlinear periodic boundary value problems ^[30], HIV nonlinear system ^[31], SIR nonlinear system of dengue fever ^[32] and alternative approach based on fuzzy-neuro methods to solve linear and nonlinear optimization problems ^[33,34]. These submissions enhance the worth of the stochastic solvers to authenticate the convergence and precision of the suggested ANN-PSO-IPA. The general form of the MS-TKNS is written as ^[35]:

$\left\{\begin{array}{l}\frac{{d}^{3}U}{d{\chi }^{3}}+\frac{2{\alpha }_{1}}{\chi }\frac{{d}^{2}U}{d{\chi }^{2}}+\frac{P(P-1)}{{\chi }^{2}}\frac{dU}{d\chi }+{H}_{1}\left(\chi \right){F}_{1}(U, V) = {G}_{1}\left(\chi \right), \\ \frac{{d}^{3}V}{d{\chi }^{3}}+\frac{2{\alpha }_{2}}{\chi }\frac{{d}^{2}V}{d{\chi }^{2}}+\frac{Q(Q-1)}{{\chi }^{2}}\frac{dV}{d\chi }+{H}_{2}\left(\chi \right){F}_{2}(U, V) = {G}_{2}\left(\chi \right), \\ U\left(0\right) = A, \frac{dU\left(0\right)}{d\chi } = \frac{{d}^{2}U\left(0\right)}{d{\chi }^{2}} = 0, \\ V\left(0\right) = B, \frac{dV\left(0\right)}{d\chi } = \frac{{d}^{2}V\left(0\right)}{d{\chi }^{2}} = 0.\end{array}\right.$

(1)

Where ${F}_{1}$ and ${F}_{2}$ are the nonlinear functions, $P$ and $Q$ are positive constants, ${G}_{1}$ and ${G}_{2}$ are indicated as a source functions.

The purpose of this study is to present the solution of the model (1) via intelligent computing ANN-PSO-IPA. The contributions of the paper are as follows:

● A novel neuro-swarm computing intelligent heuristics ANN-PSO-IPA is accessible for multi-singular nonlinear third order EF-SDEs.

● The overlapping outcomes of the proposed ANN-PSO-IPA with the exact outcomes for three examples of MS-TKNS enhance the exactness, consistency and convergence.

● Authorization of the precise performance is validated via statistical remark using the ANN-PSO-IPA based on Theil's Inequality Coefficient (TIC), Root Mean Square Error (RMSE), Variance Account For (VAF), Semi Interquartile (SI) Range.

● Beside practically precise continuous outcomes on whole input training intermission, an easy implementable process, simplicity in perception, stability and robustness are other well-intentioned announcements for the designed neuro-swarm intelligent computing approach.

The remaining structure of the current study is given as: Section 2 indicates the design methodology through PSO-IPA. The mathematical form of performance measures can be found in section 3. Section 4 shows the numerical results of the designed ANN-PSO-IPA. In the final section, final submissions and future guidance are provided.

2. Methodology

The design of ANN-PSO-IPA for MS-TKNS is presented in two steps, given as:

Step 1: An error based objective function is accessible by using the mean square error sense.

Step 2: The learning procedure of the structures is presented using the hybrid of PSO-IPA.

2.1. ANN modeling

The ANNs are famous to solve the various applications in different domain of applied science and engineering. The proposed outcomes are denoted by $U\left(\chi \right)$ and $V\left(\chi \right)$ , while $\frac{{d}^{n}U}{d{\chi }^{n}}$ and $\frac{{d}^{n}V}{d{\chi }^{n}}$ show the n^th derivative, mathematically given as:

$\left[\widehat{U}\left(\chi \right), \widehat{V}\left(\chi \right)\right] = \left[\sum\limits _{i = 1}^{m}{a}_{U, i}Z({w}_{U, i}\chi +{\phi }_{U, i}), \sum\limits _{i = 1}^{m}{a}_{V, i}Z({w}_{V, i}\chi +{\phi }_{V, i})\right],\\ \left[\frac{{d}^{n}\widehat{U}}{d{\chi }^{n}}, \frac{{d}^{n}\widehat{V}}{d{\chi }^{n}}\right] = \left[\sum\limits _{i = 1}^{m}{a}_{U, i}\frac{{d}^{n}}{d{\chi }^{n}}Z({w}_{U, i}\chi +{\phi }_{U, i}), \sum\limits _{i = 1}^{m}{a}_{V, i}\frac{{d}^{n}}{d{\chi }^{n}}Z({w}_{V, i}\chi +{\phi }_{V, i})\right].$

(2)

Where m shows the neurons and n is the derivative order. The unknown weight vectors are $a, w$ and $\phi$ . $W = [{W}_{U}, {W}_{V}]$ , for ${W}_{U} = [{\alpha }_{U}, {w}_{U}, {\phi }_{U}]$ and ${W}_{V} = [{\alpha }_{V}, {w}_{V}, {\phi }_{V}]$ . The components of the weight vector are given as:

${a}_{U} = [{a}_{U, 1}, {a}_{U, 2}, {a}_{U, 3}, ..., {a}_{U, m}], {a}_{V} = [{a}_{V, 1}, {a}_{V, 2}, {a}_{V, 3}, ..., {a}_{V, m}], \\ {w}_{U} = [{w}_{U, 1}, {w}_{U, 2}, {w}_{U, 3}, ..., {w}_{U, m}], {w}_{V} = [{w}_{V, 1}, {w}_{V, 2}, {w}_{V, 3}, ..., {w}_{V, m}], \\ {\phi }_{U} = [{\phi }_{U, 1}, {\phi }_{U, 2}, {\phi }_{U, 3}, ..., {\phi }_{U, m}], {\phi }_{V} = [{\phi }_{V, 1}, {\phi }_{V, 2}, {\phi }_{V, 3}, ..., {\phi }_{V, m}].$

The log-sigmoid $Z\left(\chi \right) = \frac{1}{\left(1+{e}^{-\chi }\right)}$ is used as an activation function. The updated form of the system (2) using the approximate results of $\widehat{U}\left(\chi \right)$ and $\widehat{V}\left(\chi \right)$ are written as:

$\left[\widehat{U}\left(\chi \right), \widehat{V}\left(\chi \right)\right] = \left[\sum\limits _{i = 1}^{m}\frac{{a}_{U, i}}{1+{e}^{-\left({w}_{U, i}x+{\phi }_{U, i}\right)}}, \sum\limits _{i = 1}^{m}\frac{{a}_{V, i}}{1+{e}^{-\left({w}_{V, i}x+{\phi }_{V, i}\right)}}\right],\\ \left[\frac{d\widehat{U}}{d\chi }, \frac{d\widehat{V}}{d\chi }\right] = \left[\sum\limits _{i = 1}^{m}\frac{{a}_{U, i}{w}_{U, i}{e}^{-\left({w}_{U, i}x+{\phi }_{U, i}\right)}}{{\left(1+{e}^{-\left({w}_{U, i}x+{\phi }_{U, i}\right)}\right)}^{2}}, \sum\limits _{i = 1}^{m}\frac{{a}_{V, i}{w}_{V, i}{e}^{-\left({w}_{V, i}x+{\phi }_{V, i}\right)}}{{\left(1+{e}^{-\left({w}_{V, i}x+{\phi }_{V, i}\right)}\right)}^{2}}\right], \\ \left[\frac{{d}^{2}\widehat{U}}{d{\chi }^{2}}, \frac{{d}^{2}\widehat{V}}{d{\chi }^{2}}\right] = \left[\begin{array}{l}\sum\limits _{i = 1}^{m}{a}_{U, i}{w}_{U, i}^{2}\left\{\frac{2{e}^{-2\left({w}_{U, i}x+{\phi }_{U, i}\right)}}{{\left(1+{e}^{-\left({w}_{U, i}x+{\phi }_{U, i}\right)}\right)}^{3}}-\frac{{e}^{-\left({w}_{U, i}x+{\phi }_{U, i}\right)}}{{\left(1+{e}^{-\left({w}_{U, i}x+{\phi }_{U, i}\right)}\right)}^{2}}\right\}, \\ \sum\limits _{i = 1}^{m}{a}_{V, i}{w}_{V, i}^{2}\left\{\frac{2{e}^{-2\left({w}_{V, i}x+{\phi }_{V, i}\right)}}{{\left(1+{e}^{-\left({w}_{V, i}x+{\phi }_{V, i}\right)}\right)}^{3}}-\frac{{e}^{-\left({w}_{V, i}x+{\phi }_{V, i}\right)}}{{\left(1+{e}^{-\left({w}_{V, i}x+{\phi }_{V, i}\right)}\right)}^{2}}\right\}\end{array}\right], \\ \left[\frac{{d}^{3}\widehat{U}}{d{\chi }^{3}}, \frac{{d}^{3}\widehat{V}}{d{\chi }^{3}}\right] = \left[\begin{array}{l}\sum\limits _{i = 1}^{m}{a}_{U, i}{w}_{U, i}^{3}\left\{\frac{6{e}^{-3\left({w}_{U, i}x+{\phi }_{U, i}\right)}}{{\left(1+{e}^{-\left({w}_{U, i}x+{\phi }_{U, i}\right)}\right)}^{4}}-\frac{6{e}^{-2\left({w}_{U, i}x+{\phi }_{U, i}\right)}}{{\left(1+{e}^{-\left({w}_{U, i}x+{\phi }_{U, i}\right)}\right)}^{3}}+\frac{{e}^{-\left({w}_{U, i}x+{\phi }_{U, i}\right)}}{{\left(1+{e}^{-\left({w}_{U, i}x+{\phi }_{U, i}\right)}\right)}^{2}}\right\}, \\ \sum\limits _{i = 1}^{m}{a}_{V, i}{w}_{V, i}^{3}\left\{\frac{6{e}^{-3\left({w}_{V, i}x+{\phi }_{V, i}\right)}}{{\left(1+{e}^{-\left({w}_{V, i}x+{\phi }_{V, i}\right)}\right)}^{4}}-\frac{6{e}^{-2\left({w}_{V, i}x+{\phi }_{V, i}\right)}}{{\left(1+{e}^{-\left({w}_{V, i}x+{\phi }_{V, i}\right)}\right)}^{3}}+\frac{{e}^{-\left({w}_{V, i}x+{\phi }_{V, i}\right)}}{{\left(1+{e}^{-\left({w}_{V, i}x+{\phi }_{V, i}\right)}\right)}^{2}}\right\}\end{array}\right]$

(3)

The error based objective formulation is written as:

$E = {E}_{1}+{E}_{2}+{E}_{3} ,$

(4)

${E}_{1} = \frac{1}{N}\sum\limits _{m = 1}^{N}{\left({\chi }_{m}^{2}\frac{{d}^{3}\widehat{U}}{d{\chi }^{3}}+2{\alpha }_{1}{\chi }_{m}\frac{{d}^{2}\widehat{U}}{d{\chi }^{2}}+P(P-1)\frac{d\widehat{U}}{d\chi }+{\chi }_{m}^{2}{H}_{1}{F}_{1}(\widehat{U}, \widehat{V})-{\chi }_{m}^{2}{G}_{1}\right)}^{2},$

(5)

${E}_{2} = \frac{1}{N}\sum\limits _{m = 1}^{N}{\left({\chi }_{m}^{2}\frac{{d}^{3}\widehat{V}}{d{\chi }^{3}}+2{\alpha }_{2}{\chi }_{m}\frac{{d}^{2}\widehat{V}}{d{\chi }^{2}}+Q(Q-1)\frac{d\widehat{V}}{d\chi }+{\chi }_{m}^{2}{H}_{2}{F}_{2}(\widehat{U}, \widehat{V})-{\chi }_{m}^{2}{G}_{2}\right)}^{2},$

(6)

${E}_{3} = \frac{1}{6}\left({\left(\widehat{U}-A\right)}^{2}+{\left({\widehat{U}}_{1}'\right)}^{2}+{\left({\widehat{U}}_{1}''\right)}^{2}+{\left(\widehat{V}-B\right)}^{2}+{\left({\widehat{V}}'\right)}^{2}+{\left({\widehat{V}}''\right)}^{2}\right) \text{ at } \chi = 0 .$

(7)

Where $N = \frac{1}{h}, {x}_{m} = mh$ . The Objective functions ${E}_{1}$ and ${E}_{2}$ are associated with the system of differential equations and ${E}_{3}$ is the corresponding initial conditions.

2.2. Optimization procedure

The optimization is performed to solve the MS-TKNS using the hybrid framework of PSO- IPA.

PSO is an effective research method that has widely used as an alternative optimization of genetic algorithms that were discovered by Kennedy and Eberhart ^[36]. In the theory of search space, a single candidate result of decision variables in the optimization procedure is called a particle and set of these particle formulated a swarms. For the refinement of optimization variables in standard PSO utilized iterative process of optimizing based on local ${P}_{LB}^{\rho -1}$ and global ${P}_{GB}^{\rho -1}$ best position of the particle in the swarm. The mathematical relations of position X_i along with the velocity V_i in PSO are given, respectively, as follows:

${X}_{i}^{\rho } = {X}_{i}^{\rho -1}+{V}_{i}^{\rho -1},$

(8)

${V}_{i}^{\rho } = \omega {V}_{i}^{\rho -1}+{\eta }_{1}({P}_{LB}^{\rho -1}-{X}_{i}^{\rho -1}){r}_{1}+{\eta }_{2}({P}_{GB}^{\rho -1}-{X}_{i}^{\rho -1}){r}_{2}$

(9)

where $\rho$ represent the current flight index, the inertia vector is denoted by $\omega$ varying between 0 and 1, ${\eta }_{1}$ and ${\eta }_{2}$ indicate the cognitive and social accelerations, respectively, while, r₁ and r₂ are vectors form with pseudo real number between 0 and 1. Further information regrading PSO can be seen in ^[37], while few recent applications address by PSO include parameter estimation ^[38], nonlinear electric circuits ^[39], optimize performance of induction generator ^[40], optimization of permanent magnets synchronous motor ^[41] and systems of equations based physical models ^[42].

The quickly converges performance of PSO is attained by the process of hybridization with the appropriate local search approach by taking the PSO best values as an initial weight. Consequently, in the presented study, an effective local search scheme based on interior-point (IPA) is exploited for rapid fine-tuning of the results by the PSO algorithm. The hybrid of PSO-IPA train the ANNs as well as fundamental parameter setting for both PSO and IPA are tabulated in Table 1. Recently, IPA is used to power flow optimization incorporating security constraints ^[43], multistage nonlinear nonconvex problems ^[44], image processing ^[45] and multi-fractional order doubly singular model ^[46]. The hybrid of PSO-IPA train the decision variables of ANNs as per procedure and settings tabulated in Table 1.

Table 1. Detailed pseudo code of PSO-IPA to solve the nonlinear third order EF-SDEs.

Start of PSO
Step-1: Initialization: Generate the primary swarm randomlyand amend the parameters of {PSO} and {optimoptions} routine.
Step-2: Fitness Evaluation: Scrutinize the {fitness value} for each particle in Eq (4).
Step-3: Ranking: Rank to each particle of the least standards of the {fitness function}
Step-4: Stopping Standards: Stop, if any of the below form achieved
● Selected flights
● Fitness level
When accomplished the above values, then go to Step-5
Step-5: Renewal: The Eqs (8) and (9) are used for the position and velocity.
Step-6: Improvement: Repeat the above steps 02–06, until the entire flights are attained.
Step-7: Storage: The attained best fitness values is stored and elect as the global best particle.
End of PSO
Start the process of PSO-IPA
Inputs: Use the global best values
Output: W_PSO-IPS are the PSO-IPA's best values
Initialize: Take {global best values} as a {start point}
Termination: Terminate the process, when {Fitness =

$E$ = 10⁻¹⁸}, {TolX = 10⁻²⁰}^, {TolCon = TolFun = 10⁻²¹}, {MaxFunEvals = 260000} and {Generation = 1500}.
While: [Stop]
Fitness Evaluation: The Eq (4) is used for the fitness value

$E$
Adjustments: Invoke the routine {fmincon} for the IP algorithm to adjust the values of the weight vector.
Store to fitness values using the basic form of the weight vector
Store: W_PSO-IPS values, final adaptive weight values, function count, fitness, time and generations for the present run.
End of the PSO-IPA

| Show Table

DownLoad: CSV

3. Performance indices

The current study is associated to present the statistical measures for solving the MS-TKNS. Therefore, three performances based on Theil's inequality coefficient (TIC) mean absolute deviation (MAD) and Variance Account For (VAF) and their global variables are Global TIC (G.TIC), Global MAD (G.MAD) and Global EVAF (G.EVAF) are applied. The mathematical descriptions of these statistical operators are provided as:

$\left[\text{TI}{\text{C}}_{U}\text{, TI}{\text{C}}_{V}\right] = \left[\frac{\sqrt{\frac{1}{n}{\sum _{i = 1}^{n}\left(U\left({\chi }_{i}\right)-\widehat{U}\left({\chi }_{i}\right)\right)}^{2}}}{\left(\sqrt{\frac{1}{n}\sum _{i = 1}^{n}{U}^{2}\left({\chi }_{i}\right)}+\sqrt{\frac{1}{n}\sum _{i = 1}^{n}{\widehat{U}}^{2}\left({\chi }_{i}\right)}\right)}, \frac{\sqrt{\frac{1}{n}{\sum _{i = 1}^{n}\left(V\left({\chi }_{i}\right)-\widehat{V}\left({\chi }_{i}\right)\right)}^{2}}}{\left(\sqrt{\frac{1}{n}\sum _{i = 1}^{n}{V}^{2}\left({\chi }_{i}\right)}+\sqrt{\frac{1}{n}\sum _{i = 1}^{n}{\widehat{V}}^{2}\left({\chi }_{i}\right)}\right)}\right],$

(10)

$\left[RMS{E}_{U}, RMS{E}_{V}\right] = \left[\sqrt{\frac{1}{n}\sum\limits _{i = 1}^{n}{\left({U}_{i}-{\widehat{U}}_{i}\right)}^{2}}, \sqrt{\frac{1}{n}\sum\limits _{i = 1}^{n}{\left({V}_{i}-{\widehat{V}}_{i}\right)}^{2}}\right],$

(11)

$\left\{\begin{array}{l}[VA{F}_{U}, VA{F}_{V}] = \left[\left(1-\frac{\mathit{var}\left(U\left({\chi }_{i}\right)-\widehat{U}\left({\chi }_{i}\right)\right)}{\mathit{var}\left(U\left({\chi }_{i}\right)\right)}\right)*100, \left(1-\frac{\mathit{var}\left(V\left({\chi }_{i}\right)-\widehat{V}\left({\chi }_{i}\right)\right)}{\mathit{var}\left(V\left({\chi }_{i}\right)\right)}\right)*100\right]\\ \left[EVA{F}_{U}, EVA{F}_{V}\right] = \left[\left|100-VA{F}_{U}\right|, \left|100-VA{F}_{V}\right|\right].\end{array}\right.$

(12)

4. Results and discussion

In this section, the detail discussion to solve three variants of the MS-TKNS is presented.

Problem 1: Consider the MS-TKNS is:

$\left\{ \begin{array}{l} \frac{{{d^3}U}}{{d{\chi ^3}}} + \frac{2}{\chi }\frac{{{d^2}U}}{{d{\chi ^2}}} + \frac{3}{\chi }\frac{{dU}}{{d\chi }} - {U^2}V = 17{e^{{\chi ^3}}} + 9\chi {e^{{\chi ^3}}} + 72{\chi ^3}{e^{{\chi ^3}}} + 27{\chi ^6}{e^{{\chi ^3}}}, \\ \frac{{{d^3}V}}{{d{\chi ^3}}} + \frac{2}{\chi }\frac{{{d^2}V}}{{d{\chi ^2}}} + \frac{3}{\chi }\frac{{dV}}{{d\chi }} - U{V^2} = - 19{e^{{\chi ^3}}} + 9\chi {e^{ - {\chi ^3}}} + 72{\chi ^3}{e^{ - {\chi ^3}}} - 27{\chi ^6}{e^{ - {\chi ^3}}}, \\ U(0) = 1, \, \, \frac{{dU(0)}}{{d\chi }} = \frac{{{d^2}U(0)}}{{d{\chi ^2}}} = 0, \, \, \\ V(0) = 1, \, \, \frac{{dV(0)}}{{d\chi }} = \frac{{{d^2}V(0)}}{{d{\chi ^2}}} = 0. \end{array} \right.$

(13)

The exact/true solutions of the above Eq (13) are [ ${e}^{{\chi }^{3}}, {e}^{-{\chi }^{3}}$ ] and the fitness function becomes as:

$E = \frac{1}{N}\sum\limits _{m = 0}^{N}\left(\begin{array}{l}{\left({\chi }_{m}\frac{{d}^{3}\widehat{U}}{d{\chi }^{3}}+2\frac{{d}^{2}\widehat{U}}{d{\chi }^{2}}+3\frac{d\widehat{U}}{d\chi }-{\chi }_{m}{\widehat{U}}^{2}\widehat{V}-{\chi }_{m}{G}_{1}\right)}^{2}+\\ {\left({\chi }_{m}\frac{{d}^{3}\widehat{V}}{d{\chi }^{3}}+2\frac{{d}^{2}\widehat{V}}{d{\chi }^{2}}+3\frac{d\widehat{V}}{d\chi }-{\chi }_{m}\widehat{U}{\widehat{V}}^{2}-{\chi }_{m}{G}_{2}\right)}^{2}\end{array}\right)\\ +\frac{1}{6}\left(\begin{array}{l}{\left(\widehat{U}-1\right)}^{2}+{\left(\frac{d\widehat{U}}{d\chi }\right)}^{2}+{\left(\frac{{d}^{2}\widehat{U}}{d{\chi }^{2}}\right)}^{2}\\ +{\left(\widehat{V}-1\right)}^{2}+{\left(\frac{d\widehat{V}}{d\chi }\right)}^{2}+{\left(\frac{{d}^{2}\widehat{V}}{d{\chi }^{2}}\right)}^{2}\end{array}\right).$

(14)

Problem 2: Consider the MS-TKNS is:

$\left\{\begin{array}{l}\frac{{d}^{3}U}{d{\chi }^{3}}+\frac{2}{\chi }\frac{{d}^{2}U}{d{\chi }^{2}}+\frac{3}{\chi }\frac{dU}{d\chi }+U{V}^{3} = 4+{e}^{-3{\chi }^{3}}+4{e}^{-2{\chi }^{3}}+6{e}^{-{\chi }^{3}}+19{e}^{{\chi }^{3}}+9\chi {e}^{{\chi }^{3}}\\ +72{\chi }^{3}{e}^{{\chi }^{3}}+27{\chi }^{6}{e}^{{\chi }^{3}}, \\ \frac{{d}^{3}V}{d{\chi }^{3}}+\frac{2}{\chi }\frac{{d}^{2}V}{d{\chi }^{2}}+\frac{3}{\chi }\frac{dV}{d\chi }+{U}^{3}V = 4-17{e}^{-{\chi }^{3}}+4{e}^{2{\chi }^{3}}+6{e}^{{\chi }^{3}}+3{e}^{3{\chi }^{3}}-9\chi {e}^{-{\chi }^{3}}\\ +72{\chi }^{3}{e}^{-{\chi }^{3}}-27{\chi }^{6}{e}^{-{\chi }^{3}}, \\ U\left(0\right) = 2, \frac{dU\left(0\right)}{d\chi } = \frac{{d}^{2}U\left(0\right)}{d{\chi }^{2}} = 0, \\ V\left(0\right) = 2, \frac{dV\left(0\right)}{d\chi } = \frac{{d}^{2}V\left(0\right)}{d{\chi }^{2}} = 0.\end{array}\right.$

(15)

The true solutions of the above equation are [ $1+{e}^{{\chi }^{3}}, 1+{e}^{-{\chi }^{3}}$ ] and the objective function becomes as:

$E = \frac{1}{N}\sum\limits _{m = 0}^{N}\left(\begin{array}{l}{\left({\chi }_{m}\frac{{d}^{3}\widehat{U}}{d{\chi }^{3}}+2\frac{{d}^{2}\widehat{U}}{d{\chi }^{2}}+3\frac{d\widehat{U}}{d\chi }+{\chi }_{m}\widehat{U}{\widehat{V}}^{3}-{\chi }_{m}{G}_{1}\right)}^{2}+\\ {\left({\chi }_{m}\frac{{d}^{3}\widehat{V}}{d{\chi }^{3}}+2\frac{{d}^{2}\widehat{V}}{d{\chi }^{2}}+3\frac{d\widehat{V}}{d\chi }+{\chi }_{m}{\widehat{U}}^{3}\widehat{V}-{\chi }_{m}{G}_{2}\right)}^{2}\end{array}\right)+\frac{1}{6}\left(\begin{array}{l}{\left(\widehat{U}-2\right)}^{2}+{\left(\frac{d\widehat{U}}{d\chi }\right)}^{2}+{\left(\frac{{d}^{2}\widehat{U}}{d{\chi }^{2}}\right)}^{2}\\ +{\left(\widehat{V}-2\right)}^{2}+{\left(\frac{d\widehat{V}}{d\chi }\right)}^{2}+{\left(\frac{{d}^{2}\widehat{V}}{d{\chi }^{2}}\right)}^{2}\end{array}\right).$

(16)

Problem 3: Consider the MS-TKNS is:

$\left\{\begin{array}{l}\frac{{d}^{3}U}{d{\chi }^{3}}+\frac{6}{\chi }\frac{{d}^{2}U}{d{\chi }^{2}}-\frac{2}{\chi }\frac{dU}{d\chi }+(1-U)V = 42-6\chi -{\chi }^{3}+{\chi }^{6}, \\ \frac{{d}^{3}V}{d{\chi }^{3}}+\frac{6}{\chi }\frac{{d}^{2}V}{d{\chi }^{2}}-\frac{2}{\chi }\frac{dV}{d\chi }-U(1-V) = -42+6\chi -{\chi }^{3}-{\chi }^{6}, \\ U\left(0\right) = 1, \frac{dU\left(0\right)}{d\chi } = \frac{{d}^{2}U\left(0\right)}{d{\chi }^{2}} = 0, \\ V\left(0\right) = 1, \frac{dV\left(0\right)}{d\chi } = \frac{{d}^{2}V\left(0\right)}{d{\chi }^{2}} = 0.\end{array}\right.$

(17)

The true solutions of the Eq (17) are [ $1+{\chi }^{3}, 1-{\chi }^{3}$ ] and the fitness function becomes as:

$E = \frac{1}{N}\sum\limits _{m = 0}^{N}\left(\begin{array}{l}{\left({\chi }_{m}\frac{{d}^{3}\widehat{U}}{d{\chi }^{3}}+6\frac{{d}^{2}\widehat{U}}{d{\chi }^{2}}-2\frac{d\widehat{U}}{d\chi }+{\chi }_{m}(1-\widehat{U})\widehat{V}-{\chi }_{m}{G}_{1}\right)}^{2}+\\ {\left({\chi }_{m}\frac{{d}^{3}\widehat{V}}{d{\chi }^{3}}+6\frac{{d}^{2}\widehat{V}}{d{\chi }^{2}}-2\frac{d\widehat{V}}{d\chi }+{\chi }_{m}\widehat{U}(1-\widehat{V})-{\chi }_{m}{G}_{2}\right)}^{2}\end{array}\right)\\ +\frac{1}{6}\left(\begin{array}{l}{\left(\widehat{U}-1\right)}^{2}+{\left(\frac{d\widehat{U}}{d\chi }\right)}^{2}+{\left(\frac{{d}^{2}\widehat{U}}{d{\chi }^{2}}\right)}^{2}\\ +{\left(\widehat{V}-1\right)}^{2}+{\left(\frac{d\widehat{V}}{d\chi }\right)}^{2}+{\left(\frac{{d}^{2}\widehat{V}}{d{\chi }^{2}}\right)}^{2}\end{array}\right).$

(18)

In order to find the proposed solutions of the Problems 1, 2 and 3 based on the MS-TKNS by using the proposed solver ANN-PSO-IPA for 40 multiple trials to achieve the adaptable parameters. The plots of the weight sets are shown in for $U\left(\chi \right)$ and $V\left(\chi \right)$ , respectively. These weights are the decision variables of ANNs as presented in equations 3 such that the fitness functions in (14), (16) and (18) for respective problems 1, 2 and 3 are optimized with PSO-IPA, i.e., initially for global search efficacy of PSO and fine tune with IPA for rapid local search. These sets of weights are applied in first equation of set (3) to find approximate solutions to the three problems and the mathematical form are given as:

${\widehat{U}}_{P-1} = \frac{-6.786}{1+{e}^{-(-11.644\chi +\text{15.35})}}+\frac{4.539}{1+{e}^{-(-1.651\chi +\text{2.200})}}\\ +\frac{-0.558}{1+{e}^{-(-8.143\chi +9.243)}}+...+\frac{4.377}{1+{e}^{-( 0.019\chi +\text{2.086})}} \ \ \ \ ,$

(19)

${\widehat{U}}_{P-2} = \frac{1.008}{1+{e}^{-(1.230\chi +0\text{.015})}}+\frac{4.209}{1+{e}^{-(12.45\chi -15\text{.681})}}\\ +\frac{0.336}{1+{e}^{-(-6.716\chi +6.309)}}+...-\frac{0.329}{1+{e}^{-( -1.124\chi +3\text{.324})}} \ \ \ \ ,$

(20)

${\widehat{U}}_{P-3} = \frac{-0.017}{1+{e}^{-(4.530\chi -3.801)}}+\frac{4.682}{1+{e}^{-(2.529\chi +2.999)}}\\ +\frac{5.098}{1+{e}^{-(0.715\chi +0.913)}}+...+\frac{3.159}{1+{e}^{-( -2.099\chi +8\text{.587})}} \ \ \ \ ,$

(21)

${\widehat{V}}_{P-1} = \frac{5.049}{1+{e}^{-(-2.457\chi +\text{2.070})}}+\frac{0.068}{1+{e}^{-(0.380\chi +1.115)}}\\ +\frac{-0.041}{1+{e}^{-(-0.464\chi +0.215)}}+...+\frac{1.790}{1+{e}^{-( 2.540\chi -1.143)}} \ \ \ \ ,$

(22)

${\widehat{V}}_{P-2} = \frac{-1.550}{1+{e}^{-(0.418\chi -1.586)}}-\frac{0.692}{1+{e}^{-(4.495\chi +2.403)}}\\ +\frac{1.550}{1+{e}^{-(-3.258\chi +2.387)}}+...-\frac{0.253}{1+{e}^{-( -0.32\chi -0.418)}} \ \ \ \ ,$

(23)

${\widehat{V}}_{P-3} = \frac{2.237}{1+{e}^{-(0.454\chi -0.035)}}-\frac{0.825}{1+{e}^{-(1.564\chi +0.541)}}\\ -\frac{0.249}{1+{e}^{-(1.204\chi -0.301)}}+...-\frac{0.771}{1+{e}^{-( -\mathrm{1,792}\chi +0.440)}} \ \ \ \ ,$

(24)

Figure 1. Best weight sets and result comparisons for Problems 1, 2 and 3.

DownLoad: Full-Size Img PowerPoint

The optimization of the MS-TKNS is performed for the problems 1, 2 and 3 using the proposed solver ANN-PSO-IPA for 40 independent trials. Set of weights and results comparison are plotted graphically in . It is specified that the exact and proposed solutions overlapped for both the indexes $\widehat{U}\left(\chi \right)$ and $\widehat{V}\left(\chi \right)$ of the problems 1, 2 and 3. This exact matches of the outcomes shows the correctness of the proposed methodology ANN-PSO-IPA. In order to calculate the comparison of the numerical results, the plots of the absolute error (AE) are drawn in , for $\widehat{U}\left(\chi \right)$ and $\widehat{V}\left(\chi \right)$ . One can observe that most of the AE values of problems 1–3 for $\widehat{U}\left(\chi \right)$ lie in the range of 10⁻⁶ to 10⁻⁷, 10⁻⁴ to 10⁻⁶ and 10⁻⁶ to 10⁻⁸, while, for $\widehat{V}\left(\chi \right)$ , these values lie around 10⁻⁵ to 10⁻⁶, 10⁻³ to 10⁻⁴ and 10⁻⁶ to 10⁻⁷. The plots of the performance measures through fitness, RMSE, TIC and EVAF are drawn in the , for $\widehat{U}\left(\chi \right)$ and $\widehat{V}\left(\chi \right)$ . It is seen that the fitness values lie around to 10⁻⁰⁶ to 10_,⁻⁰⁸ for problems 1 and 3, while the fitness values for Problem 2 are close to 10⁻⁰⁸. The RMSE values of $\widehat{U}\left(\chi \right)$ and $\widehat{V}\left(\chi \right)$ for problem 1 and 3 are close to 10⁻⁶ to 10⁻⁸, while for Problem 2, the RMSE lie 10⁻⁴–10⁻⁶ for $\widehat{U}\left(\chi \right)$ and 10⁻²–10^–4 for $\widehat{V}\left(\chi \right)$ . The TIC performance lie around 10⁻⁶–10⁻⁸ for both indexes of all the Problems. The EVAF values for problem 1 and 3 lie 10⁻¹⁰–10⁻¹² for both the indexes, while it lies around 10⁻⁸–10⁻¹⁰ for problem 2.

Figure 2. Absolute error and performance indices for Problems 1, 2 and 3.

DownLoad: Full-Size Img PowerPoint

The convergence measures for the Problems 1–3 based on the MS-TKNS using the Fitness, histograms and boxplots for 10 numbers of neurons are provided in Figure 3. It is shown that most of the fitness values lie around 10⁻⁴–10⁻⁶ for Problem 1 and 3, while for Problem 3 these values lie around 10⁻⁶–10⁻⁸. The convergence of both the indexes of all the problems for RMSE, TIC and EVAF is provided in Figures 4–9. Most of the values for both the indexes of all the problems lie in good ranges.

Figure 3. Convergence procedures for the Problems 1, 2 and 3 based on the MS-TKNS using the Fitness, histograms and boxplots for 10 neurons.

DownLoad: Full-Size Img PowerPoint

Figure 4. Convergence investigations of

$\widehat{U}\left(\chi \right)$ for the Problems 1, 2 and 3 based on the MS-TKNS using the Fitness, histograms and boxplots for 10 neurons.

DownLoad: Full-Size Img PowerPoint

Figure 5. Convergence investigations of

$\widehat{V}\left(\chi \right)$ for the Problems 1, 2 and 3 MS-TKNS using the Fitness, histograms and boxplots for 10 neurons.

DownLoad: Full-Size Img PowerPoint

Figure 6. Convergence investigations of

$\widehat{U}\left(\chi \right)$ for the Problems 1, 2 and 3 based on the MS-TKNS for the TIC values.

DownLoad: Full-Size Img PowerPoint

Figure 7. Convergence investigations of

$\widehat{V}\left(\chi \right)$ for the Problems 1, 2 and 3 based on the MS-TKNS for the TIC values.

DownLoad: Full-Size Img PowerPoint

Figure 8. Convergence measures of

$\widehat{U}\left(\chi \right)$ for the Problems 1, 2 and 3 based on the MS-TKNS using the EVAF values.

DownLoad: Full-Size Img PowerPoint

Figure 9. Convergence measures of

$\widehat{V}\left(\chi \right)$ for the Problems 1, 2 and 3 based on the MS-TKNS for the EVAF, values.

DownLoad: Full-Size Img PowerPoint

For more accuracy and precision, statistical indices are performed based on minimum (Min), standard deviation (SD), mean, SI range and Median. SI Range is one half of the difference of Q₃ = 75% data, i.e., 3^rd quartile and Q₁ = 25% data, i.e., 1^st quartile is calculated for 40 trials of ANN-PSO-IPA to solve three different problems of MS-TKNS. These statistical based outcomes for Problems 1–3 are tabulated in and for the indexes $\widehat{U}\left(\chi \right)$ and $\widehat{V}\left(\chi \right)$ . It is observed that most of the $\widehat{U}\left(\chi \right)$ and $\widehat{V}\left(\chi \right)$ values for Problems 1–3 lie in the best ranges. The global performance G.FIT, G.RMSE, G.TIC and G.EVAF of $\widehat{U}\left(\chi \right)$ and $\widehat{V}\left(\chi \right)$ for Problems 1, 2 and 3 are tabulated in . These performances of the global values for Problems 1, 2 and 3 for 40 independent trials are provided. The magnitude (Mag) and Median values for all Problems using the indexes $\widehat{U}\left(\chi \right)$ and $\widehat{V}\left(\chi \right)$ proved very good results based on the statistical global operators.

Table 2. The statistics results of

$\widehat{U}\left(\chi \right)$ for each problem of the MS-TKNS using the designed ANN-PSO-IPA.

	Mode	The solution of $\widehat{U}\left(\chi \right)$ for Problems 1–3
	Mode	0	0.1	0.2	0.3	0.4	0.5	0.6	0.7	0.8	0.9	1
P-1	Min	3×10⁻⁷	1×10⁻⁷	6×10⁻⁷	7×10⁻⁷	5×10⁻⁷	1×10⁻⁷	8×10⁻⁷	8×10⁻⁷	4×10⁻⁷	1×10⁻⁷	5×10⁻⁷
	Mean	2×10⁻⁵	2×10⁻⁵	2×10⁻⁵	3×10⁻⁵	3×10⁻⁵	4×10⁻⁵	5×10⁻⁵	5×10⁻⁵	6×10⁻⁵	6×10⁻⁵	6×10⁻⁵
	SD	4×10⁻⁵	4×10⁻⁵	4×10⁻⁵	4×10⁻⁵	4×10⁻⁵	5×10⁻⁵	5×10⁻⁵	6×10⁻⁵	6×10⁻⁵	6×10⁻⁵	6×10⁻⁵
	Median	7×10⁻⁶	6×10⁻⁶	9×10⁻⁶	1×10⁻⁵	1×10⁻⁵	2×10⁻⁵	2×10⁻⁵	3×10⁻⁵	3×10⁻⁵	3×10⁻⁵	3×10⁻⁵
	SIR	1×10⁻⁵	1×10⁻⁵	1×10⁻⁵	1×10⁻⁵	2×10⁻⁵	3×10⁻⁵	3×10⁻⁵	3×10⁻⁵	3×10⁻⁵	3×10⁻⁵	4×10⁻⁵
P-2	Min	7×10⁻⁶	8×10⁻⁶	7×10⁻⁶	1×10⁻⁵	3×10⁻⁶	2×10⁻⁶	1×10⁻⁵	1×10⁻⁶	1×10⁻⁶	1×10⁻⁵	2×10⁻⁵
	Mean	3×10⁻³	3×10⁻³	3×10⁻³	3×10⁻³	4×10⁻³	6×10⁻³	1×10⁻²	1×10⁻²	2×10⁻²	4×10⁻²	5×10⁻²
	SD	1×10⁻²	1×10⁻²	1×10⁻²	1×10⁻²	1×10⁻²	2×10⁻²	5×10⁻²	9×10⁻²	1×10⁻¹	2×10⁻¹	3×10⁻¹
	Median	3×10⁻⁴	3×10⁻⁴	3×10⁻⁴	2×10⁻⁴	2×10⁻⁴	2×10⁻⁴	2×10⁻⁴	2×10⁻⁴	2×10⁻⁴	2×10⁻⁴	3×10⁻⁴
	SIR	2×10⁻⁴	2×10⁻⁴	2×10⁻⁴	2×10⁻⁴	2×10⁻⁴	2×10⁻⁴	2×10⁻⁴	2×10⁻⁴	2×10⁻⁴	2×10⁻⁴	2×10⁻⁴
P-3	Min	1×10⁻⁷	1×10⁻⁷	7×10⁻⁷	1×10⁻⁷	2×10⁻⁷	1×10⁻⁹	3×10⁻⁷	1×10⁻⁷	8×10⁻⁷	2×10⁻⁷	1×10⁻⁷
	Mean	8×10⁻⁶	9×10⁻⁶	1×10⁻⁵	1×10⁻⁵	1×10⁻⁵	1×10⁻⁵	2×10⁻⁵	2×10⁻⁵	2×10⁻⁵	3×10⁻⁵	3×10⁻⁵
	SD	1×10⁻⁵	1×10⁻⁵	1×10⁻⁵	2×10⁻⁵	2×10⁻⁵	3×10⁻⁵	3×10⁻⁵	4×10⁻⁵	4×10⁻⁵	5×10⁻⁵	6×10⁻⁵
	Median	3×10⁻⁶	3×10⁻⁶	3×10⁻⁶	5×10⁻⁶	6×10⁻⁶	7×10⁻⁶	8×10⁻⁶	1×10⁻⁵	1×10⁻⁵	1×10⁻⁵	1×10⁻⁵
	SIR	4×10⁻⁶	4×10⁻⁶	5×10⁻⁶	6×10⁻⁶	5×10⁻⁶	6×10⁻⁶	8×10⁻⁶	9×10⁻⁶	1×10⁻⁵	1×10⁻⁵	1×10⁻⁵

| Show Table

DownLoad: CSV

Table 3. The statistics results of

$\widehat{V}\left(\chi \right)$ for each problem of the MS-TKNS using the designed ANN-PSO-IPA.

	Mode	The solution of $\widehat{U}\left(\chi \right)$ for Problems 1–3
	Mode	0	0.1	0.2	0.3	0.4	0.5	0.6	0.7	0.8	0.9	1
P-1	Min	7×10⁻⁷	4×10⁻⁷	1×10⁻⁷	1×10⁻⁶	1×10⁻⁷	1×10⁻⁶	8×10⁻⁷	2×10⁻⁶	8×10⁻⁷	3×10⁻⁶	5×10⁻⁶
	Mean	3×10⁻⁵	3×10⁻⁵	3×10⁻⁵	4×10⁻⁵	5×10⁻⁵	5×10⁻⁵	6×10⁻⁵	7×10⁻⁵	7×10⁻⁵	8×10⁻⁵	8×10⁻⁵
	SD	5×10⁻⁵	4×10⁻⁵	5×10⁻⁵	5×10⁻⁵	6×10⁻⁵	7×10⁻⁵	8×10⁻⁵	9×10⁻⁵	9×10⁻⁵	1×10⁻⁴	1×10⁻⁴
	Median	1×10⁻⁵	1×10⁻⁵	1×10⁻⁵	2×10⁻⁵	2×10⁻⁵	3×10⁻⁵	4×10⁻⁵	4×10⁻⁵	5×10⁻⁵	5×10⁻⁵	5×10⁻⁵
	SIR	1×10⁻⁵	1×10⁻⁵	1×10⁻⁵	1×10⁻⁵	1×10⁻⁵	2×10⁻⁵	2×10⁻⁵	2×10⁻⁵	2×10⁻⁵	2×10⁻⁵	2×10⁻⁵
P-2	Min	1×10⁻⁵	4×10⁻⁶	1×10⁻⁵	3×10⁻⁶	2×10⁻⁵	3×10⁻⁶	1×10⁻⁷	3×10⁻⁵	6×10⁻⁴	4×10⁻³	1×10⁻²
	Mean	1×10⁻³	5×10⁻³	1×10⁻²	2×10⁻²	3×10⁻²	4×10⁻²	4×10⁻²	5×10⁻²	5×10⁻²	6×10⁻²	9×10⁻²
	SD	3×10⁻³	2×10⁻²	9×10⁻²	1×10⁻¹	2×10⁻¹	2×10⁻¹	2×10⁻¹	3×10⁻¹	3×10⁻¹	3×10⁻¹	3×10⁻¹
	Median	2×10⁻⁴	1×10⁻⁴	2×10⁻⁴	2×10⁻⁴	2×10⁻⁴	2×10⁻⁴	2×10⁻⁴	3×10⁻⁴	2×10⁻³	9×10⁻³	3×10⁻²
	SIR	2×10⁻⁴	2×10⁻⁴	2×10⁻⁴	2×10⁻⁴	2×10⁻⁴	1×10⁻⁴	1×10⁻⁴	2×10⁻⁴	2×10⁻⁴	2×10⁻⁴	3×10⁻⁴
P-3	Min	2×10⁻⁷	3×10⁻⁷	6×10⁻⁷	6×10⁻⁷	3×10⁻⁷	2×10⁻⁷	1×10⁻⁷	1×10⁻⁷	2×10⁻⁷	1×10⁻⁷	4×10⁻⁷
	Mean	9×10⁻⁶	8×10⁻⁶	9×10⁻⁶	1×10⁻⁵	1×10⁻⁵	1×10⁻⁵	1×10⁻⁵	1×10⁻⁵	2×10⁻⁵	2×10⁻⁵	2×10⁻⁵
	SD	1×10⁻⁵	1×10⁻⁵	1×10⁻⁵	1×10⁻⁵	1×10⁻⁵	2×10⁻⁵	2×10⁻⁵	3×10⁻⁵	3×10⁻⁵	4×10⁻⁵	5×10⁻⁵
	Median	3×10⁻⁶	2×10⁻⁶	3×10⁻⁶	4×10⁻⁶	5×10⁻⁶	6×10⁻⁶	8×10⁻⁶	8×10⁻⁶	9×10⁻⁶	1×10⁻⁵	1×10⁻⁵
	SIR	5×10⁻⁶	5×10⁻⁶	3×10⁻⁶	3×10⁻⁶	4×10⁻⁶	6×10⁻⁶	6×10⁻⁶	7×10⁻⁶	8×10⁻⁶	8×10⁻⁶	1×10⁻⁵

| Show Table

DownLoad: CSV

Table 4. Global performance of

$\widehat{U}\left(\chi \right)$ and

$\widehat{V}\left(\chi \right)$ for Problems 1, 2 and 3.

Index	Example	G.FIT		G.RMSE		G.TIC		G.EVAF
Index	Example	Mag	Median	Mag	Median	Mag	Median	Mag	Median
$\widehat{U}\left(\chi \right)$	1	6×10^-6	3×10^-6	4×10⁻⁵	2×10⁻⁵	7×10⁻⁶	4×10⁻⁶	2×10⁻⁹	5×10⁻¹⁰
	2	2×10⁻⁴	4×10⁻⁵	2×10⁻²	2×10⁻⁴	7×10⁻⁶	4×10⁻⁶	4×10⁻²	2×10⁻⁷
	3	1×10⁻⁶	3×10⁻⁷	2×10⁻⁵	9×10⁻⁶	1×10⁻⁵	6×10⁻⁶	4×10⁻⁹	1×10⁻¹⁰
$\widehat{V}\left(\chi \right)$	1	6×10⁻⁶	3×10⁻⁶	6×10⁻⁵	4×10⁻⁵	1×10⁻⁵	8×10⁻⁶	3×10⁻⁷	5×10⁻⁹
	2	2×10⁻⁴	4×10⁻⁵	5×10⁻²	1×10⁻²	1×10⁻⁵	7×10⁻⁶	3×10⁻¹	2×10⁻³
	3	1×10⁻⁶	3×10⁻⁷	1×10⁻⁵	8×10⁻⁶	1×10⁻⁵	1×10⁻⁵	3×10⁻⁹	9×10⁻¹¹

| Show Table

DownLoad: CSV

5. Conclusions

In this research study, a stable, reliable and accurate numerical ANN-PSO-IPA is accessible to solve the multi-singular nonlinear third kind of Emden-Fowler system by using the ANN strength with continuous mapping. A fitness function of these networks is optimized for the global and local search capabilities of particle swarm optimization and interior-point algorithm, respectively. The proposed ANN-PSO-IPA is broadly applied to solve three different variants of the multi-singular nonlinear third kind of Emden-Fowler system. The precise and accurate performance is observed for ANN-PSO-IPA based on AE with consistent precision around 5 to 8 decimal places of precision for all three problems of the multi-singular nonlinear third kind of Emden-Fowler system. Statistical interpretations in terms of Min, Mean, SD, SI ranges and Median are performed to validate the convergence, robustness and accuracy of the proposed ANN-PSO-IPA for solving the multi-singular nonlinear third kind of Emden-Fowler system based Eqs 1–3.

In the future, new stochastic solvers based on ANN optimized with evolutionary/swarming paradigm looks proficient to solve nonlinear biological systems ^{[47,48,49,50]}, fluid dynamics models ^{[51,52,53,54,55,56]} and fractional models ^{[57,58,59,60]}. Additionally, the different ANNs structure exploiting variety of activation functions should be implemented to solve the MS-TKNS for improved performance.

Acknowledgments

This paper has been partially supported by Ministerio de Ciencia, Innovacion y Universidades grant number PGC2018-0971-B-100 and Fundacion Seneca de la Region de Murcia grant number 20783/PI/18.

Conflict of interest

All the authors of the manuscript declared that there are no potential conflicts of interest.

References

[1]	W. Mothes, N. M. Sherer, J. Jin, P. Zhong, Virus cell-to-cell transmission, J. Virol., 17 (2010), 8360-8368.
[2]	S. Iwami, J. Takeuchi, S. Nakaoka, F. Mammano, F. Clavel, H. Inaba, et al., Cell-to-cell infection by HIV contributes over half of virus infection, eLife, 23 (2015), e08150.
[3]	L. Agosto, P. Uchil, W. Mothes, Hiv cell-to-cell transmission: Effects on pathogenesis and antiretroviral therapy, Trends Microbiol., 23 (2015), 289-295.
[4]	K. Hattaf, Spatiotemporal dynamics of a generalized viral infection model with distributed delays and CTL immune response, Computation, 7 (2019), 21.
[5]	X. Ren, Y. Tian, L. Liu, X. Liu, A reaction-diffusion within-host HIV model with cell-to-cell transmission, J. Math. Biol., 76 (2018), 1831-1872.
[6]	J. Wang, J. Yang, T. Kuniya, Dynamics of a PDE viral infection model incorporating cell-to-cell transmission, J. Math. Anal. Appl., 444 (2016), 1542-1564.
[7]	R. Cui, K. Lam, Y. Lou, Dynamics and asymptotic profiles of steady states of an epidemic model in advective environments, J. Differ. Eq., 263 (2017), 2343-2373.
[8]	P. Magal, G. Webb, Y. Wu, On a vector-host epidemic model with spatial structure, Nonlinearity, 31 (2018), 5589-5614.
[9]	Y. Wu, X. Zou, Dynamics and profiles of a diffusive host-pathogen system with distinct dispersal rates, J. Differ. Eq., 264 (2018), 4989-5024.
[10]	A. Pazy, Semigroups of Linear Operators and Applications to Partial Differential Equations, Springer, New York, 1983.
[11]	H.R. Thieme, Spectral bound and reproduction number for infinite-dimensional population structure and time heterogeneity, SIAM J. Appl. Math., 70 (2009), 188-211.
[12]	G. F. Webb, Theory of Nonlinear Age-Dependent Population Dynamics, CRC Press, New York, 1985.

This article has been cited by:

1.	Muhammad Umar, Muhammad Asif Zahoor Raja, Zulqurnain Sabir, Qasem Al-Mdallal, A computational framework to solve the nonlinear dengue fever SIR system, 2022, 25, 1025-5842, 1821, 10.1080/10255842.2022.2039640
2.	Wangxun Yu, Aiju Wang, Zhiguo Qu, New Solutions of Nonlinear Equations Constrained by Bounded Variables under Edge Computation, 2022, 2022, 1530-8677, 1, 10.1155/2022/9855672
3.	Zulqurnain Sabir, Muhammad Umar, Muhammad Asif Zahoor Raja, Haci Mehmet Baskonus, Wei Gao, Designing of Morlet wavelet as a neural network for a novel prevention category in the HIV system, 2022, 15, 1793-5245, 10.1142/S1793524522500127
4.	Zulqurnain Sabir, Mohamed R. Ali, Irwan Fathurrochman, Muhammad Asif Zahoor Raja, R. Sadat, Dumitru Baleanu, Dynamics of multi-point singular fifth-order Lane–Emden system with neuro-evolution heuristics, 2022, 13, 1868-6478, 795, 10.1007/s12530-021-09413-1
5.	Sakda Noinang, Zulqurnain Sabir, Gilder Cieza Altamirano, Muhammad Asif Zahoor Raja, Manuel Jes鷖 S鄋chez-Chero, Mar韆-Ver髇ica Seminario-Morales, Wajaree Weera, Thongchai Botmart, Swarming Computational Techniques for the Influenza Disease System, 2022, 73, 1546-2226, 4851, 10.32604/cmc.2022.029437
6.	Zulqurnain Sabir, Muhammad Asif Zahoor Raja, Juan L. G. Guirao, Tareq Saeed, Swarm Intelligence Procedures Using Meyer Wavelets as a Neural Network for the Novel Fractional Order Pantograph Singular System, 2021, 5, 2504-3110, 277, 10.3390/fractalfract5040277
7.	Tianqi Yu, Lei Liu, Yan-Jun Liu, Observer-based adaptive fuzzy output feedback control for functional constraint systems with dead-zone input, 2022, 20, 1551-0018, 2628, 10.3934/mbe.2023123
8.	Zhanrong Guan, Weak signal detection method based on nonlinear differential equations, 2024, 24, 14727978, 1207, 10.3233/JCM-247329
9.	Zulqurnain Sabir, Tareq Saeed, Juan L. G. Guirao, Juan M. Sánchez, Adrián Valverde, A Swarming Meyer Wavelet Computing Approach to Solve the Transport System of Goods, 2023, 12, 2075-1680, 456, 10.3390/axioms12050456
10.	Keliang Luo, Advancing Smart Sensor Networks and Carbon-Based Biosensors Through Artificial Intelligence: A Deep Learning Approach to Optoelectronic Device Innovation, 2025, 13, 2169-3536, 86083, 10.1109/ACCESS.2025.3567561
11.	Fei Ren, Chao Ren, Optimizing physical education strategies through circular intuitionistic Fuzzy Bonferroni based school policy formulation, 2025, 15, 2045-2322, 10.1038/s41598-025-03363-3

Reader Comments

Your name:*

Email:*
© 2020 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)

通讯作者: 陈斌, bchen63@163.com

1.
沈阳化工大学材料科学与工程学院沈阳 110142

Mathematical Biosciences and Engineering

4.4

Metrics

Article views(4093) PDF downloads(393) Cited by(6)

Preview PDF

Download XML

Export Citation

Article outline

Show full outline

Mathematical Biosciences and Engineering

Global stability analysis of a viral infection model in a critical case

Related Papers:

Abstract

1. Introduction

2. Methodology

2.1. ANN modeling

2.2. Optimization procedure

3. Performance indices

4. Results and discussion

5. Conclusions

Acknowledgments

Conflict of interest

References

This article has been cited by:

Reader Comments

通讯作者: 陈斌, bchen63@163.com

Metrics

Other Articles By Authors

Catalog

Abstract

1. Introduction

2. Methodology

2.1. ANN modeling

2.2. Optimization procedure

3. Performance indices

4. Results and discussion

5. Conclusions

Acknowledgments

Conflict of interest

References

Mathematical Biosciences and Engineering

Global stability analysis of a viral infection model in a critical case

Related Papers:

Abstract

1. Introduction

2. Methodology

2.1. ANN modeling

2.2. Optimization procedure

3. Performance indices

4. Results and discussion

5. Conclusions

Acknowledgments

Conflict of interest

References

This article has been cited by:

Reader Comments

通讯作者: 陈斌, bchen63@163.com

Metrics

Other Articles By Authors

Related pages

Tools

Export File

Citation

Format

Content

Catalog

Abstract

1. Introduction

2. Methodology

2.1. ANN modeling

2.2. Optimization procedure

3. Performance indices

4. Results and discussion

5. Conclusions

Acknowledgments

Conflict of interest

References