Next Article in Journal
Estimation of Mechanical Power Output Employing Deep Learning on Inertial Measurement Data in Roller Ski Skating
Next Article in Special Issue
Using Explainable Machine Learning to Improve Intensive Care Unit Alarm Systems
Previous Article in Journal
Mapping Relation between Contour Error Components of Crankshaft Pin Journal and Axis Position Control Error of Oscillating Grinding Machine
Previous Article in Special Issue
Medical Images Encryption Based on Adaptive-Robust Multi-Mode Synchronization of Chen Hyper-Chaotic Systems
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Artificial Neural Networks to Solve the Singular Model with Neumann–Robin, Dirichlet and Neumann Boundary Conditions

1
Faculty of Computing and Informatics, Universiti Malaysia Sabah, Jalan UMS, Kota Kinabalu Sabah 88400, Malaysia
2
Department of Mathematics and Statistics, Hazara University, Mansehra 21120, Pakistan
3
Future Technology Research Center, National Yunlin University of Science and Technology, 123 University Road, Section 3, Douliou, Yunlin 64002, Taiwan
4
PPGEE, Federal University of Piauí (UFPI), Teresina 64049-550, Brazil
5
Covilhã Delegation, Instituto de Telecomunicações, 6201-001 Covilhã, Portugal
6
GRC Department, Faculty of Applied Studies, Jeddah, King Abdulaziz University, Jeddah 21589, Saudi Arabia
7
NCRA Condition Monitoring Systems Lab, Mehran University of Engineering and Technology, Jamshoro 76020, Pakistan
8
Department of Electronics and Communication Engineering, JECRC University Jaipur, Rajasthan 303905, India
*
Author to whom correspondence should be addressed.
Sensors 2021, 21(19), 6498; https://doi.org/10.3390/s21196498
Submission received: 21 April 2021 / Revised: 12 May 2021 / Accepted: 14 May 2021 / Published: 29 September 2021
(This article belongs to the Special Issue Explainable AI in Medical Sensors)

Abstract

:
The aim of this work is to solve the case study singular model involving the Neumann–Robin, Dirichlet, and Neumann boundary conditions using a novel computing framework that is based on the artificial neural network (ANN), global search genetic algorithm (GA), and local search sequential quadratic programming method (SQPM), i.e., ANN-GA-SQPM. The inspiration to present this numerical framework comes through the objective of introducing a reliable structure that associates the operative ANNs features using the optimization procedures of soft computing to deal with such stimulating systems. Four different problems that are based on the singular equations involving Neumann–Robin, Dirichlet, and Neumann boundary conditions have been occupied to scrutinize the robustness, stability, and proficiency of the designed ANN-GA-SQPM. The proposed results through ANN-GA-SQPM have been compared with the exact results to check the efficiency of the scheme through the statistical performances for taking fifty independent trials. Moreover, the study of the neuron analysis based on three and 15 neurons is also performed to check the authenticity of the proposed ANN-GA-SQPM.

1. Introduction

The singular nonlinear models that are governed with the Emden–Fowler types of equations are considered to be a hot topic for the researchers due to their massive applications in population evolution, relativistic or fluid mechanics, chemical reactor systems, and pattern structure [1,2,3,4,5]. The EF is one of the popular models due to the singular point its the origin. Some well-known applications of the singular models are the electromagnetic structure, oscillating magnetic fields, stellar structure, catalytic diffusion reactions, continuous isotropic media, isothermal gas spheres, and classical mechanics [6,7,8,9,10,11,12]. The typical form of the EF model, which is basically a second order singular differential equation, is given as [13,14,15,16]:
d 2 z d Ω 2 + χ Ω d z d Ω + h ( Ω ) g ( z ) = u ( Ω ) , z ( 0 ) = a , d z ( 0 ) d Ω = 0 .
The model (1) takes the form of Lane–Emden equation for h ( Ω ) = 1, given as:
d 2 z d Ω 2 + χ Ω d z d Ω + g ( z ) = u ( Ω ) , z ( 0 ) = a , d z ( 0 ) d Ω = 0 .
where χ represents the shape factor, u ( Ω ) is the forcing function, h ( Ω ) and g ( z ) are the functions of Ω and z, respectively. The purpose of this research work is to solve the singular equations involving Neumann–Robin, Dirichlet, and Neumann boundary conditions using a novel computing framework that is based on the artificial neural network (ANN), global search genetic algorithm (GA), and local search sequential quadratic programming method (SQPM), i.e., ANN-GA-SQPM. The general form of the singular models, along with the Neumann–Robin, Dirichlet, and Neumann boundary conditions, is written as [17]:
d 2 z d Ω 2 + χ Ω d z d Ω + g ( z ) = u ( Ω ) , z ( 0 ) = A 1 , z ( 1 ) = B 1 , d z ( 0 ) d Ω = A 2 , d z ( 1 ) d Ω = B 2 , d z ( 0 ) d Ω = 0 , A 3 z ( 1 ) + B 3 d z ( 1 ) d Ω = C 1
where A 1 , B 1 , A 2 , B 2 , A 3 , B 3 , and C 1 are the real constants. For the numerical outcomes of the singular system, a variety of applications have been investigated in the references [18,19,20]. All of these stated approaches have their separate perks and importance, as well as confines and disadvantages. Alongside these conventional schemes, the stochastic design of the computational numerical heuristic or swarming techniques appears to be capable and proficient in integrating the singular systems by operating the universal approximation capability of ANNs along with local/global techniques. The nonlinear prey-predator system [21], nonlinear singular Thomas–Fermi system [22], SITR nonlinear based COVID model [23,24], periodic singular differential system [25,26], dengue fever model [27], multi-singular systems [28], HIV infection model [29], nonlinear singular functional differential model [30,31], heat conduction system in human head [32], and mosquito dispersal model [33] are some recent submissions of the stochastic solvers. When considering these influences, the authors are inspired to present the solution of the singular models involving the Neumann–Robin, Dirichlet, and Neumann boundary conditions using the ANN-GA-SQPM. Some novel characteristics of the ANN-GA-SQPM are briefly provided as:
  • A pioneering framework using the integrated computational ANN-GA-SQPM is provided to solve the singular model involving the Neumann–Robin, Dirichlet, and Neumann boundary conditions.
  • The performance of the computational ANN-GA-SQPM is observed using a small and large number of neurons.
  • The matching of the results that were obtained by the proposed computational ANN-GA-SQPM with the exact solutions authenticate the value in terms of convergence and precision.
  • The absolute error (AE) is found in good measure for each problem of the singular model.
  • The verification of the ANN-GA-SQPM is authorized from the statistical exploration on multiple executions for 10 neurons based on the performance of Variance Account For (VAF), Nash Sutcliffe Efficiency (NSE), and Theil’s Inequality Coefficient (TIC).
  • Besides the equitable precise solutions of the system, the easy understanding, smooth operations, robustness, and comprehensive stability are other valued merits.

2. Methodology: ANN-GA-SQPM

The design structure is presented in two phases to solve the singular model. An error function is introduced and the hybridization procedures of GA-SQPM is provided.

2.1. ANNs Modeling

In order to solve the singular models involving the Neumann–Robin, Dirichlet, and Neumann boundary conditions that accumulated with feed-forward ANNs, z ^ ( Ω ) is the continuous mapping form of the solution together with the log-sigmoid function Q ( Ω ) = 1 / 1 + exp ( Ω ) given as:
z ^ ( Ω ) = i = 1 k r i Q ( w i χ + s i ) = i = 1 k r i 1 + e ( w i χ + s i ) 1 , d z ^ d χ = i = 1 k r i d d χ Q ( w i χ + s i ) = i = 1 k r i w i e ( w i χ + s i ) 1 + e ( w i χ + s i ) 2 , d 2 z ^ d χ 2 = i = 1 k r i d 2 d χ 2 Q ( w i χ + s i ) = i = 1 k r i w i 2 2 e 2 ( w i χ + s i ) 1 + e ( w i χ + s i ) 3 e ( w i χ + s i ) 1 + e ( w i χ + s i ) 2 ,
where r = [ r 1 , r 2 , r 3 , r k ] , w = [ w 1 , w 2 , w 3 , , w k ] , and   s = [ s 1 , s 2 , s 3 , , s k ] are the unknown weights. An error function for solving the singular model (3) is given as:
ξ F i t = ξ F i t 1 + ξ F i t 2 ,
where the construction of the error functions ξ F i t 1 and ξ F i t 2 is on the basis of the singular model and the Neumann–Robin, Dirichlet, and Neumann boundary conditions, respectively, given as:
ξ F i t 1 = 1 N k = 1 N d 2 z ^ k d Ω k 2 + χ Ω k d z ^ k d Ω k + g ( z ^ k ) u k , 0 Ω k 1 ,
ξ F i t 2 = 1 2 z ^ 0 A 1 2 + 1 2 z ^ N B 1 2 ,
ξ F i t 3 = 1 2 z ^ 0 A 2 2 + 1 2 z ^ N B 2 2 ,
ξ F i t 4 = 1 2 z ^ 0 2 + 1 2 A 3 z ^ N B 3 z ^ N C 1 2 ,
where N h = 1 , z ^ k = z ( χ k ) , g ( z ^ k ) = g ( χ k ) , u k = u ( χ k ) and z k = k h .

2.2. Optimization Process: GA-SQPM

The optimization performance of the ANNs is accomplished through the hybridization procedures of GA-SQPM.
GAs are applied to solve both constrained/unconstrained optimization models based on the natural selection process. GAs are implemented frequently to transform the population of individual solutions that solve an assortment of systems using the optimization procedures, where the fundamental optimization schemes fail, e.g., non-differentiable systems, stochastic models, and highly nonlinear systems. GA is applied in various fields of technologies, applied sciences, and engineering that work through its reproduction operators. In recent years, GA and PSO based heuristic optimization solvers have been applied in transportation planning and logistics management [34], microgrid energy management systems [35], optimization of multimodal functions [36], mobile position estimation problem [37], the vehicle routing problem in cloud implementation [38], satellite formation reconfiguration [39], and the optimization of multi-objective energy models [40].
SQPM is considered to be very important in the process of optimization. SQPM is known as a local search scheme and it has been implemented to solve constrained/unconstrained models. In recent years, SQPM has been implemented in the sizing and location of DGs [41], optimal gait based on bipedal robots through nonlinear system of predictive control [42], optimal organization of directional overcurrent communication incorporating spread generation [43], second order prediction differential system [44], central air-conditioning optimization [45], and flight vehicle management [46].
One can use the process of hybridization with the local search scheme to help overcome the sluggishness and laziness associated with controlling the global scheme. Table 1 presents the pseudocode of the GA-SQPM.

3. Results and Discussion

The current section provides details of the nonlinear singular differential model based on the Neumann–Robin, Dirichlet, and Neumann boundary conditions using the designed framework of ANN-GA-SQPM. Two problems based on nonlinear singular systems with the Dirichlet boundary condition while one problem each for Neumann–Robin and Neumann boundary conditions, respectively, are implemented in order to evaluate the performance of the designed ANN-GA-SQPM. The details of the results comparison, AE, performance measures, and weight plots, along with statistical observations, are also presented.
Problem 1: Consider the following singular Lane–Emden nonlinear model along with Dirichlet boundary conditions, which are written as:
d 2 z d Ω 2 + 0.5 Ω d z d Ω + e 2 z = 0.5 e z , z ( 0 ) = log ( 2 ) , z ( 1 ) = 0 .
The exact solution for the above equation is ln 2 1 + Ω 2 . The error function is given as:
ξ F i t = 1 N k = 1 N d 2 z ^ k d Ω k 2 + 0.5 Ω k d z ^ k d Ω k + e 2 z ^ k 0.5 e z ^ k + 1 2 z ^ 0 log ( 2 ) 2 + z ^ N 2
Optimization is performed through the hybridization of GA-SQPM to calculate the numerical representations of problem 1. The numerical outcomes are derived in Table 2, Table 3, Table 4 and Table 5 for small and large neurons three, 10, and 15, respectively, using a 0.05 step size with input [0,1]. One can observe that the proposed and exact solutions for three, 10, and 15 neurons consistently overlap the exact solutions. It is also noticed that, by taking a small number of neurons, the performance of ANN-GA-SQPM is reasonably good, but greater accuracy is observed for larger, 15 neuron-based networks. However, the complexity cost increases by increasing the number of variables/neurons in the networks.
Problem 2: Consider the following singular Lane–Emden nonlinear model along with the Dirichlet boundary conditions, which are written as:
d 2 z d Ω 2 + 0.5 Ω d z d Ω Ω 2 e z ( 16 Ω 4 e z 14 ) = 0 , z ( 0 ) = log ( 0.25 ) , z ( 1 ) = log ( 0.2 ) .
The exact solution for the above equation is ln 1 4 + Ω 4 . The error function is given as:
ξ F i t = 1 N k = 1 N d 2 z ^ k d Ω k 2 + 0.5 Ω k d z ^ k d Ω k Ω k 2 e z ^ k ( 16 Ω k 4 e z ^ k 14 ) + 1 2 z ^ 0 log ( 0.25 ) 2 + z ^ N log ( 0.2 ) 2 ,
Problem 3: Consider the following singular Lane–Emden nonlinear model along with Neumann boundary conditions is written as:
d 2 z d Ω 2 + 2 Ω d z d Ω e z ( 4 Ω 2 e z 6 ) = 0 , z ( 0 ) = 0 , z ( 1 ) = 0.4 .
The exact solution for the above equation is ln 1 4 + Ω 2 . The error function is given as:
ξ F i t = 1 N k = 1 N d 2 z ^ k d Ω k 2 + 2 Ω k d z ^ k d Ω k e z ^ k ( 4 Ω k 2 e z ^ k 6 ) + 1 2 z ^ 0 2 + z ^ N + 0.4 2 ,
Problem 4: Consider the following singular Lane–Emden nonlinear model along with Neumann–Robin boundary conditions used in the modelling of isothermal gas spheres is given as:
d 2 z d Ω 2 + 2 Ω d z d Ω + z 5 = 0 , z ( 0 ) = 0 , z ( 1 ) = 0.75 .
The exact solution the for above equation is 3 3 + Ω 2 . The error function is given as:
ξ F i t = 1 N k = 1 N d 2 z ^ k d Ω k 2 + 2 Ω k d z ^ k d Ω k + z k 5 2 + 1 2 z ^ 0 2 + z ^ N 0.75 2 ,

4. Investigation through Multiple Executions of ANN-GA-SQPM

The proposed results through ANN-GA-SQPM for fifty independent trials to accomplish the system parameter for the singular models that involve Neumann–Robin, Dirichlet, and Neumann boundary conditions are given in Equations in set (3). The best weights set is applied to designate the obtained results of the singular model, being mathematically given as:
z ^ 1 ( Ω ) = 3.7679 1 + e ( 3.012 Ω 3.285 ) + 0.0376 1 + e ( 0.388 Ω 0.334 ) + 1.1037 1 + e ( 0.173 Ω 3.2126 ) + 0.6175 1 + e (   0.222 Ω + 1.136 ) ,
z ^ 2 ( Ω ) = 2.1015 1 + e ( 4.269 Ω 13.333 ) 5.5141 1 + e ( 4.496 Ω + 11.462 ) + 8.9057 1 + e ( 7.248 Ω 12.503 ) + 6.4430 1 + e (   0.073 Ω + 1.069 ) ,
z ^ 3 ( Ω ) = 5.5275 1 + e ( 17.177 Ω + 19.987 ) 0.9686 1 + e ( 0.297 Ω + 14.586 ) 13.6035 1 + e ( 1.585 Ω + 18.439 ) + 1.4923 1 + e (   1.740 Ω 11.650 ) ,
z ^ 4 ( Ω ) = 1.4575 1 + e ( 1.003 Ω 10.567 ) + 0.2586 1 + e ( 4.234 Ω + 1.983 ) 3.32035 1 + e ( 1.679 Ω 9.102 ) + 3.943 1 + e (   10.219 Ω + 10.745 ) ,
The optimal performance is provided to solve the singular model that involves Dirichlet and Neumann boundary conditions for fifty runs. Figure 1 plots a set of best weights using 30 variables.
The statistical measures of the ANN-GA-SQPM are accessible for solving the singular model of Lane Emden type, as defined in Equations in set (3), using 30 numbers of variables or 10 neurons, which is practicable in terms of complexity and accuracy as compared to three and 15 neurons. The ANN-GA-SQPM simulations have been accompanied by 50 trials to solve all four problems of those singular models, which involve Neumann–Robin, Dirichlet, and Neumann boundary conditions. The numerical results are provided on the basis of statistical performances that are graphically depicted in Figure 2 and Figure 3.
The exact, approximate results with minimum (Min) fitness (FIT) values, i.e., the best values, approximate outcomes with maximum (Max) FIT values, the worst, along with mean outcomes, are illustrated in Figure 2a–d for each singular models, which involve Neumann–Robin, Dirichlet, and Neumann boundary conditions. Nevertheless, the RMSE values at the same inputs are derived in Figure 2f–h. One can conclude that the proposed outcomes of the ANN-GA-SQPM attained a sensible precision even in the worst case too, although there is no perceptible difference in the presentation by deviation of the singular models that involve Neumann–Robin, Dirichlet and Neumann boundary conditions.
Moreover, the performance operatives for the FIT, RMSE, ENSE, and TIC are also considered, and their statistical based outcomes are provided in Figure 3a–d for each problem of the singular model. These consequences authenticate the comparable tendencies of performance based different measures. For problem 1, the FIT, ENSE, RMSE, and EVAF best values lie around 10−08 to 10−09, 10−11 to 10−12, 10−05 to 10−06, and 10−10 to 10−11. The mean values for problem 1 are found around 10−06 to 10−07, 10−08 to 10−09, 10−04 to 10−05, and 10−07 to 10−08. The worst values even lie around 10−04 to 10−05, 10−06 to 10−07, 10−03 to 10−04, and 10−06 to 10−07. For problem 2, the FIT, ENSE, RMSE, and EVAF best values lie around 10−09 to 10−10, 10−12 to 10−14, 10−05 to 10−07, and 10−11 to 10−13. The mean values for problem 2 are found around 10−05 to 10−07, 10−09 to 10−10, 10−04 to 10−05, and 10−06 to 10−08. The worst values even lie around 10−04 to 10−05, 10−07 to 10−08, 10−03 to 10−05, and 10−06 to 10−07. For problem 3, the FIT, ENSE, RMSE, and EVAF best values lie around 10−10 to 10−11, 10−12 to 10−13, 10−06 to 10−08, and 10−12 to 10−14. The mean values for problem 3 are found around 10−05 to 10−06, 10−08 to 10−10, 10−04 to 10−05, and 10−06 to 10−07. The worst values even lie around 10−04 to 10−05, 10−06 to 10−08, 10−02 to 10−03, and 10−06 to 10−08. For problem 4, the FIT, ENSE, RMSE, and EVAF best values lie around 10−10 to 10−11, 10−13 to 10−14, 10−06 to 10−07, and 10−13 to 10−14. The mean values for problem 4 are found around 10−05 to 10−07, 10−08 to 10−09, 10−06 to 10−07, and 10−07 to 10−08. The worst values even lie around 10−04 to 10−05, 10−06 to 10−08, 10−04 to 10−05, and 10−05 to 10−07. These optimal close values for each operator enhance the worth of the propose ANN-GA-SQPM for solving the nonlinear singular Lane–Emden system.
The complexity measures of ANN-GA-SQPM are shown in terms of time, iterations, and FIT assessed during the optimization procedures of hybrid heuristics GA-SQPM and standalone GAs to adjust the network’s decision variables. The optimization performance of GAs is degraded considerably with the increase of generations and rapid convergence achieved with the SQPM procedure, but at the cost of additional computations. The outcomes through the complexity indices are provided in Table 6 for each problem of the singular model of the Lane–Emden type for GA-SQPM. The hybrid heuristics GA-SQPM take almost 20% more computation time, i.e., 402, 805, 813, and 354 time consumed for ANN-GA for problems 1, 2, 3, and 4, respectively. Moreover, the computational time taken by three, 10, and 15 based neuron networks are found around 100 ± 10, 410 ± 300, and 750 ± 450. One may observe that no perceptible difference is perceived while using these complexity operatives with a fixed neuron for ANN-GA-SQPM to solve the singular model of Lane–Emden type that approves the smooth execution of the proposed approach for different problems.

5. Performance Operators

In this section, the performances that are based on RMSE, ENSE, and EVAF are provided. The mathematical measures of these operators are given as:
RMSE = 1 n m = 1 n Ψ m Ψ ^ m 2 .
V A F = 1 var Z i ( T ) Z ^ i ( T ) var Z i ( T ) × 100 , E V A F = V A F 100 ,
NSE = 1 m = 1 n Ψ m Ψ ^ m 2 m = 1 n Ψ m Ψ ¯ m 2 , Ψ ¯ m = 1 n m = 1 n Ψ m ,
E N S E = 1 NSE

6. Conclusions

The present study is aims to design an alternate, stable, and accurate stochastic computing numerical approach to solve the singular model of Lane–Emden type that involves Neumann–Robin, Dirichlet, and Neumann boundary conditions by manipulating the ANN, local search SQPM, and GA based global search. The proposed structure of ANN-GA-SQPM is examined for different neurons/variables in the system and the presentation is acquired reasonably for all neurons based on the networks of ANN. The most reliable and accurate solutions are attained for large neurons, but the complexity increased. Statistics results through different performances for the convergence, precision, and complexity authenticate the value of the proposed ANN-GA-SQPM to solve the singular model of the Lane–Emden type of problem that involves Neumann–Robin, Dirichlet, and Neumann boundary conditions.
In the future, one may solve the fractional form of the singular models with Neumann–Robin, Dirichlet, and Neumann boundary conditions using the proposed ANN-GA-SQPM. Additionally, the memetic computing paradigm of ANN-GA-SQPM can be a good alternative to be exploited for problems involving the study of sensors [47,48,49,50,51].

Author Contributions

Formal analysis, M.A.Z.R.; Investigation, S.R.M.; Methodology, K.N.; Resources, A.A.A.I.; Software, J.J.P.C.R. and M.G.; Visualization, B.S.C.; Writing—original draft, Z.S.; Writing—review & editing, K.N. All authors have read and agreed to the published version of the manuscript.

Funding

The manuscript APC is supported by Universiti Malaysia Sabah, Jalan UMS, 88400, Kota Kinabalu, Sabah, Malaysia. This work is partially supported by FCT/MCTES through national funds and when applicable cofunded EU funds under the Project UIDB/50008/2020; and by Brazilian National Council for Scientific and Technological Development—CNPq, via Grant No. 313036/2020-9.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Wong, J.S. On the generalized Emden–Fowler equation. Siam Rev. 1975, 17, 339–360. [Google Scholar] [CrossRef]
  2. Wazwaz, A.-M. Adomian decomposition method for a reliable treatment of the Emden–Fowler equation. Appl. Math. Comput. 2005, 161, 543–560. [Google Scholar] [CrossRef]
  3. Rach, R.; Duan, J.-S.; Wazwaz, A.-M. Solving coupled Lane–Emden boundary value problems in catalytic diffusion reactions by the Adomian decomposition method. J. Math. Chem. 2014, 52, 255–267. [Google Scholar] [CrossRef]
  4. Taghavi, A.; Pearce, S.K. A solution to the Lane-Emden equation in the theory of stellar structure utilizing the Tau method. Math. Methods Appl. Sci. 2013, 36, 1240–1247. [Google Scholar] [CrossRef]
  5. Boubaker, K.; van Gorder, R.A. Application of the BPES to Lane–Emden equations governing polytropic and iso-thermal gas spheres. New Astron. 2012, 17, 565–569. [Google Scholar] [CrossRef]
  6. Hadian-Rasanan, A.H.; Rahmati, D.; Gorgin, S.; Parand, K. A single layer fractional orthogonal neural network for solving various types of Lane–Emden equation. New Astron. 2020, 75, 101307. [Google Scholar] [CrossRef]
  7. Sabir, Z.; Raja, M.A.Z.; Khalique, C.M.; Unlu, C. Neuro-evolution computing for nonlinear multi-singular system of third order Emden–Fowler equation. Math. Comput. Simul. 2021, 185, 799–812. [Google Scholar] [CrossRef]
  8. Džurina, J.; Grace, S.R.; Jadlovská, I.; Li, T. Oscillation criteria for second-order Emden–Fowler delay differential equations with a sublinear neutral term. Math. Nachr. 2020, 293, 910–922. [Google Scholar] [CrossRef]
  9. Sabir, Z.; Guirao, J.L.; Saeed, T. Solving a novel designed second order nonlinear Lane–Emden delay differential model using the heuristic techniques. Appl. Soft Comput. 2021, 102, 107105. [Google Scholar] [CrossRef]
  10. Khan, I.; Raja, M.A.Z.; Shoaib, M.; Kumam, P.; Alrabaiah, H.; Shah, Z.; Islam, S. Design of Neural Network With Levenberg-Marquardt and Bayesian Regularization Backpropagation for Solving Pantograph Delay Differential Equations. IEEE Access 2020, 8, 137918–137933. [Google Scholar] [CrossRef]
  11. Sabir, Z.; Raja, M.A.Z.; Guirao, J.L.G.; Shoaib, M. Integrated intelligent computing with neuro-swarming solver for multi-singular fourth-order nonlinear Emden–Fowler equation. Comput. Appl. Math. 2020, 39, 1–18. [Google Scholar] [CrossRef]
  12. Singh, K.; Verma, A.K.; Singh, M. Higher order Emden–Fowler type equations via uniform Haar Wavelet resolution technique. J. Comput. Appl. Math. 2020, 376, 112836. [Google Scholar]
  13. Sabir, Z.; Wahab, H.A.; Umar, M.; Sakar, M.G.; Raja, M.A.Z. Novel design of Morlet wavelet neural network for solving second order Lane–Emden equation. Math. Comput. Simul. 2020, 172, 1–14. [Google Scholar] [CrossRef]
  14. Adel, W.; Sabir, Z. Solving a new design of nonlinear second-order Lane–Emden pantograph delay differential model via Bernoulli collocation method. Eur. Phys. J. Plus 2020, 135, 1–12. [Google Scholar] [CrossRef]
  15. Abdelkawy, M.A.; Sabir, Z.; Guirao, J.L.; Saeed, T. Numerical investigations of a new singular second-order non-linear coupled functional Lane–Emden model. Open Phys. 2020, 18, 770–778. [Google Scholar] [CrossRef]
  16. Guirao, J.L.; Sabir, Z.; Saeed, T. Design and Numerical Solutions of a Novel Third-Order Nonlinear Emden–Fowler Delay Differential Model. Math. Probl. Eng. 2020, 2020, 1–9. [Google Scholar] [CrossRef]
  17. Singh, R.; Garg, H.; Guleria, V. Haar wavelet collocation method for Lane–Emden equations with Dirichlet, Neumann and Neumann–Robin boundary conditions. J. Comput. Appl. Math. 2019, 346, 150–161. [Google Scholar] [CrossRef]
  18. Sabir, Z.; Günerhan, H.; Guirao, J.L. On a new model based on third-order nonlinear multisingular functional differential equations. Math. Probl. Eng. 2020, 2020, 1–9. [Google Scholar] [CrossRef] [Green Version]
  19. Asadpour, S.; Yazdani Cherati, A.; Hosseinzadeh, H. Solving the general form of the Emden-Fowler equations with the Moving Least Squares method. J. Math. Model. 2019, 7, 231–250. [Google Scholar]
  20. Moaaz, O.; Elabbasy, E.M.; Qaraad, B. An improved approach for studying oscillation of generalized Emden–Fowler neutral differential equation. J. Inequalities Appl. 2020, 2020, 1–18. [Google Scholar] [CrossRef]
  21. Umar, M.; Sabir, Z.; Raja, M.A.Z. Intelligent computing for numerical treatment of nonlinear prey–predator models. Appl. Soft Comput. 2019, 80, 506–524. [Google Scholar] [CrossRef]
  22. Sabir, Z.; Manzar, M.A.; Raja, M.A.Z.; Sheraz, M.; Wazwaz, A.M. Neuro-heuristics for nonlinear singular Thomas-Fermi systems. Appl. Soft Comput. 2018, 65, 152–169. [Google Scholar] [CrossRef]
  23. Umar, M.; Sabir, Z.; Raja, M.A.Z.; Amin, F.; Saeed, T.; Guerrero-Sanchez, Y. Integrated neuro-swarm heuristic with interior-point for nonlinear SITR model for dynamics of novel COVID-19. Alex. Eng. J. 2021, 60, 2811–2824. [Google Scholar] [CrossRef]
  24. Umar, M.; Sabir, Z.; Raja, M.A.Z.; Shoaib, M.; Gupta, M.; Sánchez, Y.G. A Stochastic Intelligent Computing with Neuro-Evolution Heuristics for Nonlinear SITR System of Novel COVID-19 Dynamics. Symmetry 2020, 12, 1628. [Google Scholar] [CrossRef]
  25. Sabir, Z.; Raja, M.A.Z.; Guirao, J.L.; Shoaib, M. A neuro-swarming intelligence based computing for second order singular periodic nonlinear boundary value problems. Front. Phys. 2020, 8, 224. [Google Scholar] [CrossRef]
  26. Sabir, Z.; Khalique, C.M.; Raja, M.A.Z.; Baleanu, D. Evolutionary computing for nonlinear singular boundary value problems using neural network, genetic algorithm and active-set algorithm. Eur. Phys. J. Plus 2021, 136, 1–19. [Google Scholar] [CrossRef]
  27. Umar, M.; Sabir, Z.; Raja, M.A.Z.; Sánchez, Y.G. A stochastic numerical computing heuristic of SIR nonlinear model based on dengue fever. Results Phys. 2020, 19, 103585. [Google Scholar] [CrossRef]
  28. Raja, M.A.Z.; Mehmood, J.; Sabir, Z.; Nasab, A.K.; Manzar, M.A. Numerical solution of doubly singular nonlinear systems using neural networks-based integrated intelligent computing. Neural Comput. Appl. 2019, 31, 793–812. [Google Scholar] [CrossRef]
  29. Umar, M.; Sabir, Z.; Amin, F.; Guirao, J.L.G.; Raja, M.A.Z. Stochastic numerical technique for solving HIV infection model of CD4+ T cells. Eur. Phys. J. Plus 2020, 135, 1–19. [Google Scholar] [CrossRef]
  30. Sabir, Z.; Wahab, H.A.; Umar, M.; Erdoğan, F. Stochastic numerical approach for solving second order nonlinear singular functional differential equation. Appl. Math. Comput. 2019, 363, 124605. [Google Scholar] [CrossRef]
  31. Sabir, Z.; Raja, M.A.Z.; Umar, M.; Shoaib, M. Neuro-swarm intelligent computing to solve the second-order singular functional differential model. Eur. Phys. J. Plus 2020, 135, 474. [Google Scholar] [CrossRef]
  32. Raja, M.A.Z.; Umar, M.; Sabir, Z.; Khan, J.A.; Baleanu, D. A new stochastic computing paradigm for the dynamics of nonlinear singular heat conduction model of the human head. Eur. Phys. J. Plus 2018, 133, 364. [Google Scholar] [CrossRef]
  33. Umar, M.; Raja, M.A.Z.; Sabir, Z.; Alwabli, A.S.; Shoaib, M. A stochastic computational intelligent solver for numerical treatment of mosquito dispersal model in a heterogeneous environment. Eur. Phys. J. Plus 2020, 135, 1–23. [Google Scholar] [CrossRef]
  34. Arkhipov, D.I.; Wu, D.; Wu, T.; Regan, A.C. A Parallel Genetic Algorithm Framework for Transportation Planning and Logistics Management. IEEE Access 2020, 8, 106506–106515. [Google Scholar] [CrossRef]
  35. Leonori, S.; Paschero, M.; Mascioli, F.M.F.; Rizzi, A. Optimization strategies for Microgrid energy management systems by Genetic Algorithms. Appl. Soft Comput. 2020, 86, 105903. [Google Scholar] [CrossRef]
  36. Cao, Y.; Zhang, H.; Li, W.; Zhou, M.; Zhang, Y.; Chaovalitwongse, W.A. Comprehensive Learning Particle Swarm Optimization Algorithm With Local Search for Multimodal Functions. IEEE Trans. Evol. Comput. 2019, 23, 718–731. [Google Scholar] [CrossRef]
  37. Yue, Y.; Cao, L.; Hu, J.; Cai, S.; Hang, B.; Wu, H. A Novel Hybrid Location Algorithm Based on Chaotic Particle Swarm Optimization for Mobile Position Estimation. IEEE Access 2019, 7, 58541–58552. [Google Scholar] [CrossRef]
  38. Abbasi, M.; Rafiee, M.; Khosravi, M.R.; Jolfaei, A.; Menon, V.G.; Koushyar, J.M. An efficient parallel genetic algorithm solution for vehicle routing problem in cloud implementation of the intelligent transportation systems. J. Cloud Comput. 2020, 9, 6. [Google Scholar] [CrossRef]
  39. Sarno, S.; Guo, J.; D’Errico, M.; Gill, E. A guidance approach to satellite formation reconfiguration based on convex optimization and genetic algorithms. Adv. Space Res. 2020, 65, 2003–2017. [Google Scholar] [CrossRef]
  40. Salata, F.; Ciancio, V.; Dell’Olmo, J.; Golasi, I.; Palusci, O.; Coppi, M. Effects of local conditions on the multi-variable and multi-objective energy optimization of residential buildings using genetic algorithms. Appl. Energy 2020, 260, 114289. [Google Scholar] [CrossRef]
  41. Montoya, O.D.; Gil-González, W.; Grisales-Noreña, L. Relaxed convex model for optimal location and sizing of DGs in DC grids using sequential quadratic programming and random hyperplane approaches. Int. J. Electr. Power Energy Syst. 2020, 115, 105442. [Google Scholar] [CrossRef]
  42. Sun, Z.; Zhang, B.; Sun, Y.; Pang, Z.; Cheng, C. A Novel Superlinearly Convergent Trust Region-Sequential Quadratic Programming Approach for Optimal Gait of Bipedal Robots via Nonlinear Model Predictive Control. J. Intell. Robot. Syst. 2020, 100, 401–416. [Google Scholar] [CrossRef]
  43. ElSayed, S.K.; Elattar, E.E. Hybrid Harris hawks optimization with sequential quadratic programming for optimal coordination of directional overcurrent relays incorporating distributed generation. Alex. Eng. J. 2021, 60, 2421–2433. [Google Scholar] [CrossRef]
  44. Sabir, Z.; Raja, M.A.Z.; Wahab, H.A.; Shoaib, M.; Aguilar, J.G. Integrated neuro-evolution heuristic with sequential quadratic programming for second-order prediction differential models. Numer. Methods Partial Differ. Equ. 2020. [Google Scholar] [CrossRef]
  45. Xie, J.; Zhang, H.; Shen, Y.; Li, M. Energy consumption optimization of central air-conditioning based on sequential-least-square-programming. In Proceedings of the 2020 Chinese Control and Decision Conference (CCDC), Hefei, China, 22–24 August 2020; IEEE: Piscataway, NJ, USA, 2020; pp. 5147–5152. [Google Scholar]
  46. Hong, H.; Maity, A.; Holzapfel, F. Free Final-Time Constrained Sequential Quadratic Programming–Based Flight Vehicle Guidance. J. Guid. Control. Dyn. 2021, 44, 181–189. [Google Scholar] [CrossRef]
  47. Zgank, A. IoT-based bee swarm activity acoustic classification using deep neural networks. Sensors 2021, 21, 676. [Google Scholar] [CrossRef] [PubMed]
  48. Zhang, J.; Lu, C.; Wang, J.; Yue, X.G.; Lim, S.J.; Al-Makhadmeh, Z.; Tolba, A. Training convolutional neural networks with multi-size images and triplet loss for remote sensing scene classification. Sensors 2020, 20, 1188. [Google Scholar] [CrossRef] [Green Version]
  49. Francik, S.; Kurpaska, S. The Use of Artificial Neural Networks for Forecasting of Air Temperature inside a Heated Foil Tunnel. Sensors 2020, 20, 652. [Google Scholar] [CrossRef] [Green Version]
  50. Casilari, E.; Lora-Rivera, R.; García-Lagos, F. A Study on the Application of Convolutional Neural Networks to Fall Detection Evaluated with Multiple Public Datasets. Sensors 2020, 20, 1466. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  51. Ortega, S.; Halicek, M.; Fabelo, H.; Camacho, R.; Plaza, M.D.L.L.; Godtliebsen, F.; Callicó, G.M.; Fei, B. Hyper-spectral imaging for the detection of glioblastoma tumor cells in H&E slides using convolutional neural networks. Sensors 2020, 20, 1911. [Google Scholar]
Figure 1. Best weights to solve the singular model of Lane–Emden type.
Figure 1. Best weights to solve the singular model of Lane–Emden type.
Sensors 21 06498 g001
Figure 2. Comparison of results along with the AE through ANN-GA-SQPM to solve the singular model of Lane–Emden type. (ad) for the solution dynamics, while (eh) for AE.
Figure 2. Comparison of results along with the AE through ANN-GA-SQPM to solve the singular model of Lane–Emden type. (ad) for the solution dynamics, while (eh) for AE.
Sensors 21 06498 g002
Figure 3. Performance measures through ANN-GA-SQPM to solve the singular model of Lane–Emden type.
Figure 3. Performance measures through ANN-GA-SQPM to solve the singular model of Lane–Emden type.
Sensors 21 06498 g003
Table 1. Pseudocode of the process of optimization using the ANN-GA-SQPM.
Table 1. Pseudocode of the process of optimization using the ANN-GA-SQPM.
Start of GA
Inputs:
The chromosome with the same entries of the system are signified as:
W = [ r , w , s ]
Population: The chromosomes set is designated as:
r = [ r 1 , r 2 , r 3 , r k ] , w = [ w 1 , w 2 , w 3 , , w k ]   and   s = [ s 1 , s 2 , s 3 , , s k ]
P = [ W 1 , W 2 , , , W k ] t
Output: The best weights of GA are WBest-GA
Initialization
Create W that is a WBest-GA of real numbers to signify a chromosome. Initialize the W with real entries. Adjust the ‘Generation’ & ‘declarations’ values of ‘gaoptimset’ & GA routines
Fitness formulation
Accomplish the ξ F i t in P to show all W for Equations (5)–(8)
Termination
Stop the process to accomplish
•  ξ F i t = 10 −18, TolFun = 10−21, Generations = 100, ,
• TolCon = 10−22, Population Size = 270, StallGenLimit = 120
Go to [storage], when stopping standards obtains.
Ranking
Rank W of P for brilliance of ξ F i t
Storage
Save WBest-GA, iterations, ξ F i t and time for the current trials of GAs
End of GA
GA-SQPM Start
Inputs
WBest-GA is the start point
Output
WGA-SQPM represents the best values
Initialize
Adjust WGA-SQPM represents an initial input
Termination
Stop the procedure, when
ξ F i t = 1018′, generations = 1000, TolFun = 10−21,
TolX=1019′, TolCon =10−18′, MaxEvalsFun= 229,000
While [Terminate]
Fitness Calculations
Calculate ξ F i t of the present W using Equations (5–8).
Amendments
Invoke ‘fmincon’ for the SQPM. Adjust W for each generation of SQPM. Calculate
Calculate ξ F i t of updated W using Equations (5–8)
Accumulate
Store WGA-SQPM, time, ξ F i t and number of generations for the current trials of SQPM.
End of GA-SQPM Procedure
Table 2. Results comparison of Problem 1 based on ANN-GA-SQPM for three, 10, and 15 neurons, or nine, 30, and 45 variables with reference solutions.
Table 2. Results comparison of Problem 1 based on ANN-GA-SQPM for three, 10, and 15 neurons, or nine, 30, and 45 variables with reference solutions.
Ω ExactApproximate Results z ^ ( Ω )
z ^ ( Ω ) 9 Variables30 Variables45 Variables
00.693147180560.682244533460.693150901950.69314891025
0.050.690650300360.675242812410.690645711700.69064852047
0.10.683196849710.664233371830.683204634220.68319977498
0.150.670896571630.649353351930.670917173390.67090442363
0.20.653926467410.630756283970.653953966120.65393705056
0.250.632522558740.608611300290.632551612250.63253391938
0.30.606969484320.583102176090.606997640070.60698071676
0.350.577588839930.554426210420.577616237040.57759992784
0.40.544727175440.522792956790.544755201470.54473850462
0.450.508744457560.488422816900.508774397950.50875637605
0.50.470003629250.451545514230.470035849240.47001621283
0.550.428861685910.412398466340.428895481550.42887471647
0.60.385662480810.371225077430.385696460430.38567556481
0.650.340731293540.328272974060.340764009450.34074402602
0.70.294371060600.283792208510.294401584950.29438315629
0.750.246860077930.238033454720.246888282520.24687142635
0.80.198450938720.191246221830.198477365640.19846157782
0.850.149370456650.143677109910.149395829450.14938048998
0.90.099820335280.095568131370.099844936620.09982983598
0.950.049978369810.047155119770.050001686360.04998732054
100.001333754230.000021196920.00000831368
Table 3. Results comparison of Problem 2 using ANN-GA-SQPM based on three, 10, and 15 neurons or nine, 30, and 45 variables neurons with the reference solutions.
Table 3. Results comparison of Problem 2 using ANN-GA-SQPM based on three, 10, and 15 neurons or nine, 30, and 45 variables neurons with the reference solutions.
Ω ExactApproximate Results z ^ ( Ω )
z ^ ( Ω ) 9 Variables30 Variables45 Variables
0−1.38629436112−1.38629436112−1.38619926024−1.38630126484
0.05−1.38629592362−1.38629592362−1.38611885693−1.38631522992
0.1−1.38631936081−1.38631936081−1.38617352354−1.38633499755
0.15−1.38642091561−1.38642091561−1.38633380823−1.38642949431
0.2−1.38669428114−1.38669428114−1.38665342802−1.38669799402
0.25−1.38727044709−1.38727044709−1.38725068090−1.38727265740
0.3−1.38831731357−1.38831731357−1.38829571544−1.38832031881
0.35−1.39003890406−1.39003890406−1.39000152175−1.39004329283
0.4−1.39267396808−1.39267396808−1.39261717627−1.39267897863
0.45−1.39649373274−1.39649373274−1.39642230631−1.39649806188
0.5−1.40179854766−1.40179854766−1.40172202250−1.40180115494
0.55−1.40891317854−1.40891317854−1.40884175300−1.40891377590
0.6−1.41818054998−1.41818054998−1.41812154180−1.41817964531
0.65−1.42995382621−1.42995382621−1.42990946826−1.42995237211
0.7−1.44458685387−1.44458685387−1.44455392788−1.44458570299
0.75−1.46242316947−1.46242316947−1.46239459698−1.46242261605
0.8−1.48378398240−1.48378398240−1.48375199787−1.48378364332
0.85−1.50895575599−1.50895575599−1.50891569341−1.50895489962
0.9−1.53817818786−1.53817818786−1.53813127387−1.53817637114
0.95−1.57163349586−1.57163349586−1.57158645455−1.57163106694
1−1.60943791243−1.60943791243−1.60939677357−1.60943565431
Table 4. Results comparison of Problem 3 using ANN-GA-SQPM based on three, 10, and 15 neurons or nine, 30, and 45 variables with the reference solutions.
Table 4. Results comparison of Problem 3 using ANN-GA-SQPM based on three, 10, and 15 neurons or nine, 30, and 45 variables with the reference solutions.
Ω ExactApproximate Results z ^ ( Ω )
z ^ ( Ω ) 9 Variables30 Variables45 Variables
0−1.38629436112−1.48446054700−1.38630356972−1.38631710010
0.05−1.38691916589−1.48757921354−1.38692665406−1.38694119063
0.1−1.38879124132−1.49113183052−1.38879709413−1.38881050847
0.15−1.39190359988−1.49517102282−1.39190920225−1.39191976243
0.2−1.39624469197−1.49975342347−1.39625061225−1.39625835324
0.25−1.40179854766−1.50493923270−1.40180467700−1.40181070914
0.3−1.40854497005−1.51079149158−1.40855102806−1.40855665180
0.35−1.41645977545−1.51737500782−1.41646560212−1.41647178419
0.4−1.42551507427−1.52475487352−1.42552068561−1.42552789275
0.45−1.43567958630−1.53299452399−1.43568510504−1.43569335640
0.5−1.44691898294−1.54215330699−1.44692454151−1.44693355582
0.55−1.45919624910−1.55228356454−1.45920191864−1.45921127714
0.6−1.47247205736−1.56342727723−1.47247782108−1.47248710477
0.65−1.48670514705−1.57561238316−1.48671091461−1.48671979939
0.7−1.50185270175−1.58884895695−1.50185835049−1.50186665735
0.75−1.51787071891−1.60312551155−1.51787614407−1.51788384920
0.8−1.53471436624−1.61840575515−1.53471952326−1.53472673514
0.85−1.55233832039−1.63462618173−1.55234324496−1.55235015636
0.9−1.57069708412−1.65169487952−1.57070187969−1.57070870155
0.95−1.58974527909−1.66949189317−1.58975006558−1.58975694855
1−1.60943791243−1.68787136420−1.60944273339−1.60944968143
Table 5. Results comparison of Problem 4 using ANN-GA-SQPM based on three, 10 and 15 neurons or nine, 30, and 45 variables with the reference solutions.
Table 5. Results comparison of Problem 4 using ANN-GA-SQPM based on three, 10 and 15 neurons or nine, 30, and 45 variables with the reference solutions.
Ω ExactApproximate Results z ^ ( Ω )
z ^ ( Ω ) 9 Variables30 Variables45 Variables
01.000000000000.994882480941.000007838261.00000105336
0.050.999583593570.994546714290.999590830040.99958445423
0.10.998337488460.993269556100.998341753200.99833784278
0.150.996270962770.991104948640.996271650840.99627085144
0.20.993399267800.988105590220.993396896640.99339889181
0.250.989743318610.984322810120.989738982500.98974290152
0.30.985329278160.979806465610.985324218130.98532898386
0.350.980188050780.974604859370.980183348480.98018795036
0.40.974354703690.968764675250.974351097940.97435478017
0.450.967867836990.962330930740.967865652380.96786801419
0.50.960768922830.955346944350.960768092210.96076910439
0.550.953101634250.947854316260.953101791120.95310174015
0.60.944911182520.939892920850.944911796040.94491117389
0.650.936243679770.931500909480.936244203950.93624356674
0.70.927145540820.922714722360.927145550800.92714537307
0.750.917662935480.913569108300.917662226900.91766277924
0.80.907841299000.904097150990.907839931250.90784120821
0.850.897724905920.894330301130.897723175640.89772489740
0.90.887356509420.884298413200.887354846820.88735655314
0.950.876777046040.874029786140.876775832780.87677708125
10.866025403780.863551207240.866024716550.86602539000
Table 6. The complexity performance through ANN-GA-SQPM for each example of the singular model of Lane–Emden type.
Table 6. The complexity performance through ANN-GA-SQPM for each example of the singular model of Lane–Emden type.
ProblemImplementation TimeIterationsCount of Function
MinSDMinSDMinSD
1522.784672576.40890433.30000110.2983427589.960007079.39701
2861.839185412.52809472.5200073.42641029664.620004479.75041
3881.23202026.2453000467.16000104.0714730003.560006967.51609
4437.113242576.23782338.94000147.1297422032.320009767.47337
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Nisar, K.; Sabir, Z.; Asif Zahoor Raja, M.; Ag Ibrahim, A.A.; J. P. C. Rodrigues, J.; Refahy Mahmoud, S.; Chowdhry, B.S.; Gupta, M. Artificial Neural Networks to Solve the Singular Model with Neumann–Robin, Dirichlet and Neumann Boundary Conditions. Sensors 2021, 21, 6498. https://doi.org/10.3390/s21196498

AMA Style

Nisar K, Sabir Z, Asif Zahoor Raja M, Ag Ibrahim AA, J. P. C. Rodrigues J, Refahy Mahmoud S, Chowdhry BS, Gupta M. Artificial Neural Networks to Solve the Singular Model with Neumann–Robin, Dirichlet and Neumann Boundary Conditions. Sensors. 2021; 21(19):6498. https://doi.org/10.3390/s21196498

Chicago/Turabian Style

Nisar, Kashif, Zulqurnain Sabir, Muhammad Asif Zahoor Raja, Ag Asri Ag Ibrahim, Joel J. P. C. Rodrigues, Samy Refahy Mahmoud, Bhawani Shankar Chowdhry, and Manoj Gupta. 2021. "Artificial Neural Networks to Solve the Singular Model with Neumann–Robin, Dirichlet and Neumann Boundary Conditions" Sensors 21, no. 19: 6498. https://doi.org/10.3390/s21196498

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop