Optimization of tertiary building passive parameters by forecasting energy consumption based on artificial intelligence models and using ANOVA variance analysis method

Lamya Lairgi; Rachid Lagtayi; Yassir Lairgi; Abdelmajid Daya; Rabie Elotmani; Ahmed Khouya; Mohammed Touzani; Lamya Lairgi; Rachid Lagtayi; Yassir Lairgi; Abdelmajid Daya; Rabie Elotmani; Ahmed Khouya; Mohammed Touzani

doi:10.3934/energy.2023039

AIMS Energy

2023, Volume 11, Issue 5: 795-809. doi: 10.3934/energy.2023039

Previous Article Next Article

Research article

Optimization of tertiary building passive parameters by forecasting energy consumption based on artificial intelligence models and using ANOVA variance analysis method

1.
Department of Physics, Laboratory M3ER, FSTE, Errachidia, Moulay Ismail University, Meknes, Morocco
2.
Laboratory of Innovative Technologies, National School for Applied Sciences, Abdelmalek Essaadi University, B.P. 1818 Tangier, Morocco
3.
Department of Data Science, MINES School Saint Etienne France
4.
Laboratory of Sciences for Energy LabSIPE, ENSAJ, Chouaib Doukkali University, El Jadida, Morocco

Received: 14 June 2023 Revised: 24 August 2023 Accepted: 10 September 2023 Published: 18 September 2023

Energy consumption in the tertial sector is largely attributed to cooling/heating energy consumption. Thus, forecasting the building's energy consumption has become a key factor in long-term decision-making, reducing the huge energy demand and future planning. This manuscript outlines to use of the variance analysis method (ANOVA) to study the building's passive parameters' effect, such as the orientation, insulation, and its thickness plus the glazing on energy savings through the forecasting of the heating/cooling energy consumption by applying the Seasonal Auto-Regressive Integrated Moving Average (SARIMA) and the Long Short-Term Memory (LSTM) models. The presented methodology compares the predicted consumed energy of a baseline building with another efficient building which includes all the passive parameters selected by the ANOVA approach. The results show that the improvement of passive parameters leads to a reduction of heating energy consumption by 1,739,640 kWh from 2021 to 2029, which is equivalent to a monthly energy consumption of 181.2 kWh for an administrative building with an area of 415 m². While the cooling energy consumption is diminished by 893,246 kWh from 2021 to 2029, which leads to save a monthly value of 93.05 kWh. Consequently, the passive parameters optimization efficiently reduces the consumed energy and minimizes its costs. This positively impacts our environment due to the reduction of gas emissions, air and soil pollution.

Keywords:

Citation: Lamya Lairgi, Rachid Lagtayi, Yassir Lairgi, Abdelmajid Daya, Rabie Elotmani, Ahmed Khouya, Mohammed Touzani. Optimization of tertiary building passive parameters by forecasting energy consumption based on artificial intelligence models and using ANOVA variance analysis method[J]. AIMS Energy, 2023, 11(5): 795-809. doi: 10.3934/energy.2023039

Related Papers:

[1]	Bilal Akbar, Khuram Pervez Amber, Anila Kousar, Muhammad Waqar Aslam, Muhammad Anser Bashir, Muhammad Sajid Khan . Data-driven predictive models for daily electricity consumption of academic buildings. AIMS Energy, 2020, 8(5): 783-801. doi: 10.3934/energy.2020.5.783
[2]	Zhongjiao Ma, Zichun Yan, Mingfei He, Haikuan Zhao, Jialin Song . A review of the influencing factors of building energy consumption and the prediction and optimization of energy consumption. AIMS Energy, 2025, 13(1): 35-85. doi: 10.3934/energy.2025003
[3]	Kaovinath Appalasamy, R Mamat, Sudhakar Kumarasamy . Smart thermal management of photovoltaic systems: Innovative strategies. AIMS Energy, 2025, 13(2): 309-353. doi: 10.3934/energy.2025013
[4]	Feng Tian . A wind power prediction model with meteorological time delay and power characteristic. AIMS Energy, 2025, 13(3): 517-539. doi: 10.3934/energy.2025020
[5]	Xiaojing Tian, Weiqi Ye, Liang Xu, Anjian Yang, Langming Huang, Shenglong Jin . Optimization research on laminated cooling structure for gas turbines: A review. AIMS Energy, 2025, 13(2): 354-401. doi: 10.3934/energy.2025014
[6]	Emrah Gulay . Forecasting the total electricity production in South Africa: Comparative analysis to improve the predictive modelling accuracy. AIMS Energy, 2019, 7(1): 88-110. doi: 10.3934/energy.2019.1.88
[7]	Rahaman Abu, John Amakor, Rasaq Kazeem, Temilola Olugasa, Olusegun Ajide, Nosa Idusuyi, Tien-Chien Jen, Esther Akinlabi . Modeling influence of weather variables on energy consumption in an agricultural research institute in Ibadan, Nigeria. AIMS Energy, 2024, 12(1): 256-270. doi: 10.3934/energy.2024012
[8]	Yan Li, Yaheng Su, Qixin Zhao, Bala Wuda, Kaibo Qu, Lei Tang . An electricity price optimization model considering time-of-use and active distribution network efficiency improvements. AIMS Energy, 2025, 13(1): 13-34. doi: 10.3934/energy.2025002
[9]	Anbazhagan Geetha, S. Usha, J. Santhakumar, Surender Reddy Salkuti . Forecasting of energy consumption rate and battery stress under real-world traffic conditions using ANN model with different learning algorithms. AIMS Energy, 2025, 13(1): 125-146. doi: 10.3934/energy.2025005
[10]	Ghafi Kondi Akara, Benoit Hingray, Adama Diawara, Arona Diedhiou . Effect of weather on monthly electricity consumption in three coastal cities in West Africa. AIMS Energy, 2021, 9(3): 446-464. doi: 10.3934/energy.2021022

Abstract

1. Introduction

In the present paper, we study the following coupled nonlinear wave equations (see ^[1])

$\begin{align} u_{tt}-div\Big(\rho\Big(|\nabla u|^2\Big)\nabla u\Big)+|u_t|^{m-1}u_t+f_1(u, v)& = 0, \end{align}$

(1.1)

$\begin{align} v_{tt}-div\Big(\rho\Big(|\nabla v|^2\Big)\nabla v\Big)+|v_t|^{n-1}v_t+f_2(u, v)& = 0, \end{align}$

(1.2)

where $u_t = \frac{\partial u}{\partial t}, v_t = \frac{\partial v}{\partial t}$ , $m, n > 1$ , $\nabla$ is the gradient operator, div is the divergence operator and $f_i(., .):\mathbb{R}^2\rightarrow \mathbb{R}, i = 1, 2$ are known functions. For arbitrary solutions of (1.1) and (1.2), the function $\rho$ is supposed to satisfy one or the other of the two conditions:

A1). $0 < \rho^2(q^2)q^2\leq m_1\widetilde{\rho}(q^2),$

A2). $0 < \frac{1}{K_1} < \rho^{-1}(q^2)\leq K_2\widetilde{\rho}(q^2)\frac{1}{q^2},$

where $\widetilde{\rho}(q^2) = \int_0^{q^2}\rho(\zeta)d\zeta, \ q^2 = |\nabla u|^2, m_1, K_1, K_2 > 0.$

Quintanilla ^[2] imposed condition A1 and obtained the growth or decay estimates of the solution to the type Ⅲ heat conduction. The condition A1 can be satisfied easily, e.g., $\rho(q^2) = \frac{1}{\sqrt{1+q^2}}$ or $\rho(q^2) = (1+\frac{b}{p}q^2)^{p-1}, b > 0, 0 < p\leq1.$ The similar condition as A2 was also considered by many papers (see ^[3,4]). $\rho(q^2) = \sqrt{1+q^2}$ satisfies the condition A2.

In addition, we introduce a function $F(u, v)$ which is defined as $\frac{\partial F}{\partial u} = f_1(u, v), \ \frac{\partial F}{\partial v} = f_2(u, v),$ where $F(0, 0) = 0.$

The wave equations have attracted many attentions of scholars due to their wide application, and a large number of achievements have been made in the existence of solutions (see ^{[1,5,6,7,8,9,10,11,12]}). The fuzzy inference method is used to solve this problem. The algebraic formulation of fuzzy relation is studied in ^[13,14]. In this paper, we study the Phragmén-Lindelöf type alternative property of solutions of wave equations (1.1)–(1.2). It is proved that the solution of the equations either grows exponentially (polynomially) or decays exponentially (polynomially) when the space variable tends to infinite. In the case of decay, people usually expect a fast decay rate. The Phragmén-Lindelöf type alternative research on partial differential equations has lasted for a long time (see ^{[2,15,16,17,18,19,20,21,22,23]}).

It is worth emphasizing that Quintanilla ^[2] considered an exterior or cone-like region. Under some appropriate conditions, the growth/decay estimates of some parabolic problems are obtained. Inspired by ^[2], we extends the research of to the nonlinear wave model in this paper. However, different from ^[2], in addition to condition A1 and condition A2, we also consider a special condition of $\rho$ . The appropriate energy function is established, and the nonlinear differential inequality about the energy function is derived. By solving this differential inequality, the Phragmén-Lindelöf type alternative results of the solution are obtained. Our model is much more complex than ^[2], so the methods used in ^[2] can not be applied to our model directly. Finally, a nonlinear system of viscoelastic type is also studied when the system is defined in an exterior or cone-like regions and the growth or decay rates are also obtained.

2. The Phragmén-Lindelöf type alternative result under A1

Letting that $\Omega(r)$ denotes a cone-like region, i.e.,

$\Omega(r) = \Big\{\mathit{\boldsymbol{x}}\Big||\mathit{\boldsymbol{x}}|^2\geq r^2, r\geq R_0 > 0\Big\},$

and that $B(r)$ denotes the exterior surface to the sphere, i.e.,

$B(r) = \Big\{\mathit{\boldsymbol{x}}\Big||\mathit{\boldsymbol{x}}|^2 = r^2, r\geq R_0 > 0\Big\},$

Equations (1.1) and (1.2) also have the following initial-boundary conditions

$\begin{align} u(\mathit{\boldsymbol{x}}, 0) = v(\mathit{\boldsymbol{x}}, 0) = 0, \ & in\ \Omega, \end{align}$

(2.1)

$\begin{align} u(\mathit{\boldsymbol{x}}, t) = g_1(\mathit{\boldsymbol{x}}, t), \ v(\mathit{\boldsymbol{x}}, t) = g_2(\mathit{\boldsymbol{x}}, t), \ & in\ B(R_0)\times(0, \tau), \end{align}$

(2.2)

where $g_1$ and $g_2$ are positive known functions, $\textbf{x} = (x_1, x_2, x_3)$ and $\tau > 0$ .

Now, we establish an energy function

$\begin{align} \mathcal{E}(r, t)& = \int_0^t\int_{B(r)}e^{-\omega\eta}\rho\Big(|\nabla u|^2\Big)\nabla u\cdot \frac{\mathit{\boldsymbol{x}}}{r}u_\eta dSd\eta +\int_0^t\int_{B(r)}e^{-\omega\eta}\rho\Big(|\nabla v|^2\Big)\nabla v\cdot\frac{\mathit{\boldsymbol{x}}}{r}v_\eta dSd\eta\\ &\doteq \mathcal{E}_1(r, t)+\mathcal{E}_2(r, t). \end{align}$

(2.3)

Let $r_0$ be a positive constant which satisfies $r > r_0\geq R_0$ . Integrating $\mathcal{E}(z, t)$ from $r_0$ to $r$ , using the divergence theorem and Eqs (1.1) and (1.2), (2.1) and (2.2), we have

$\begin{align} \mathcal{E}(r, t)&-\mathcal{E}(r_0, t) = \int_0^t\int_{r_0}^r\int_{B(\xi)}e^{-\omega \eta}\nabla\cdot\Big[\rho\Big(|\nabla u|^2\Big)\nabla uu_\eta\Big] dsd\xi d\eta \\ &+\int_0^t\int_{r_0}^r\int_{B(\xi)}e^{-\omega \eta}\nabla\cdot\Big[\rho\Big(|\nabla v|^2\Big)\nabla vv_\eta\Big] dsd\xi d\eta \\ &+\int_0^t\int_{r_0}^r\int_{B(\xi)}e^{-\omega \eta}\Big[u_{\eta\eta}+|u_\eta|^{m+1}+f_1(u, v)\Big]u_\eta dsd\xi d\eta \\ &+\int_0^t\int_{r_0}^r\int_{B(\xi)}e^{-\omega \eta}\Big[v_{\eta\eta}+|v_\eta|^{n+1}+f_2(u, v)\Big]v_\eta dsd\xi d\eta \\ &+\frac{1}{2}\int_0^t\int_{r_0}^r\int_{B(\xi)}e^{-\omega \eta}\frac{\partial}{\partial \eta}\widetilde{\rho}\Big(|\nabla u|^2\Big) dsd\xi d\eta +\frac{1}{2}\int_0^t\int_{r_0}^r\int_{B(\xi)}e^{-\omega \eta}\frac{\partial}{\partial \eta}\widetilde{\rho}\Big(|\nabla v|^2\Big) dsd\xi d\eta \\ & = \frac{1}{2}e^{-\omega t}\int_{r_0}^r\int_{B(\xi)}\Big[|u_t|^2+|v_t|^2+\widetilde{\rho}\Big(|\nabla u|^2\Big)+\widetilde{\rho}\Big(|\nabla v|^2\Big)+2F(u, v)\Big]dsd\xi \\ &+\frac{1}{2}\omega\int_0^t\int_{r_0}^r\int_{B(\xi)}e^{-\omega\eta}\Big[|u_\eta|^2+|v_\eta|^2+\widetilde{\rho}\Big(|\nabla u|^2\Big)+\widetilde{\rho}\Big(|\nabla v|^2\Big)+2F(u, v)\Big]dSd\xi d\eta \\&+\int_0^t\int_{r_0}^r\int_{B(\xi)}e^{-\omega\eta}\Big[|u_\eta|^{m+1}+|v_\eta|^{n+1}\Big]dsd\xi d\eta, \end{align}$

(2.4)

from which it follows that

$\begin{align} \frac{\partial}{\partial r}\mathcal{E}(r, t)& = \frac{1}{2}e^{-\omega t}\int_{B(r)}\Big[|u_t|^2+|v_t|^2+\widetilde{\rho}\Big(|\nabla u|^2\Big)+\widetilde{\rho}\Big(|\nabla v|^2\Big)+2F(u, v)\Big]ds \\ &+\frac{1}{2}\omega\int_0^t\int_{B(r)}e^{-\omega\eta}\Big[|u_\eta|^2+|v_\eta|^2+\widetilde{\rho}\Big(|\nabla u|^2\Big)+\widetilde{\rho}\Big(|\nabla v|^2\Big)+2F(u, v)\Big]dS d\eta \\&+\int_0^t\int_{B(r)}e^{-\omega\eta}\Big[|u_\eta|^{m+1}+|v_\eta|^{n+1}\Big]dsd\eta, \end{align}$

(2.5)

where $\omega$ is positive constant.

Now, we show how to bound $\mathcal{E}(r, t)$ by $\frac{\partial}{\partial r}\mathcal{E}(r, t)$ . We use the Hölder inequality, the Young inequality and the condition A1 to have

$\begin{align} \Big|\mathcal{E}_1(r, t)\Big|&\leq\Big[\int_0^t\int_{B(r)}e^{-\omega\eta}\rho^2\Big(|\nabla u|^2\Big)|\nabla u|^2ds d\eta\cdot\int_0^t\int_{B(r)}e^{-\omega\eta}u_\eta^2ds d\eta\Big]^\frac{1}{2}\\ &\leq\sqrt{m_1}\Big[\int_0^t\int_{B(r)}e^{-\omega\eta}\widetilde{\rho}\Big(|\nabla u|^2\Big)ds d\eta\cdot\int_0^t\int_{B(r)}e^{-\omega\eta}u_\eta^2ds d\eta\Big]^\frac{1}{2}\\ &\leq\frac{\sqrt{m_1}}{2}\Big[\int_0^t\int_{B(r)}e^{-\omega\eta}\widetilde{\rho}\Big(|\nabla u|^2\Big)ds d\eta+\int_0^t\int_{B(r)}e^{-\omega\eta}u_\eta^2ds d\eta\Big], \end{align}$

(2.6)

and

$\begin{align} \Big|\mathcal{E}_2(r, t)\Big|&\leq\frac{\sqrt{m_1}}{2}\Big[\int_0^t\int_{B(r)}e^{-\omega\eta}\widetilde{\rho}\Big(|\nabla v|^2\Big)ds d\eta+\int_0^t\int_{B(r)}e^{-\omega\eta}v_\eta^2dsd\eta\Big]. \end{align}$

(2.7)

Inserting (2.6) and (2.7) into (2.3) and combining (2.5), we have

$\begin{align} \Big|\mathcal{E}(r, t)\Big|\leq \frac{\sqrt{m_1}}{\omega}\Big[\frac{\partial}{\partial r}\mathcal{E}(r, t)\Big]. \end{align}$

(2.8)

We consider inequality (2.8) for two cases.

I. If $\exists r_0 > R_0$ such that $\mathcal{E}(r_0, t)\geq0$ . From (2.5), we obtain $\mathcal{E}(r, t)\geq\mathcal{E}(r_0, t)\geq0, r\geq r_0$ . Therefore, from (2.8) we have

$\begin{align} \mathcal{E}(r, t)\leq \frac{\sqrt{m_1}}{\omega}\Big[\frac{\partial}{\partial r}\mathcal{E}(r, t)\Big], r\geq r_0. \end{align}$

(2.9)

Integrating (2.9) from $r_0$ to $r$ , we have

$\begin{align} \mathcal{E}(r, t)\geq\Big[\mathcal{E}(r_0, t)\Big]e^{\frac{\omega}{\sqrt{m_1}}(r-r_0)}, r\geq r_0. \end{align}$

(2.10)

Combing (2.4) and (2.10), we have

$\begin{align} \lim\limits_{r\rightarrow \infty}&\Big\{e^{-\frac{\omega}{\sqrt{m_1}}r}\Big[\frac{1}{2}e^{-\omega t}\int_{r_0}^r\int_{B(\xi)}\Big[|u_t|^2+|v_t|^2+\widetilde{\rho}\Big(|\nabla u|^2\Big)+\widetilde{\rho}\Big(|\nabla v|^2\Big)+2F(u, v)\Big]dsd\xi \\ &+\frac{1}{2}\omega\int_0^t\int_{r_0}^r\int_{B(\xi)}e^{-\omega\eta}\Big[|u_\eta|^2+|v_\eta|^2+\widetilde{\rho}\Big(|\nabla u|^2\Big)+\widetilde{\rho}\Big(|\nabla v|^2\Big)+2F(u, v)\Big]dsd\xi d\eta \\&+\int_0^t\int_{r_0}^r\int_{B(\xi)}e^{-\omega\eta}\Big[|u_\eta|^{m+1}+|v_\eta|^{n+1}\Big]dsd\xi d\eta\Big]\Big\} \\ &\geq \mathcal{E}(R_0, t)e^{-\frac{\omega}{\sqrt{m_1}}R_0}. \end{align}$

(2.11)

II. If $\forall r > R_0$ such that $\mathcal{E}(r, t) < 0$ . The inequality (2.8) can be rewritten as

$\begin{align} -\mathcal{E}(r, t)\leq \frac{\sqrt{m_1}}{\omega}\Big[\frac{\partial}{\partial r}\mathcal{E}(r, t)\Big], r\geq R_0. \end{align}$

(2.12)

Integrating (2.12) from $r_0$ to $r$ , we have

$\begin{align} -\mathcal{E}(r, t)\geq\Big[-\mathcal{E}(R_0, t)\Big]e^{-\frac{\omega}{\sqrt{m_1}}(r-R_0)}, r\geq R_0. \end{align}$

(2.13)

Inequality (2.13) shows that $\lim_{r\rightarrow \infty}\Big[-\mathcal{E}(r, t)\Big] = 0$ . Integrating (2.5) from $r$ to $\infty$ and combining (2.13), we obtain

$\begin{align} \frac{1}{2}&e^{-\omega t}\int_{r}^\infty\int_{B(\xi)}\Big[|u_t|^2+|v_t|^2+\widetilde{\rho}\Big(|\nabla u|^2\Big)+\widetilde{\rho}\Big(|\nabla v|^2\Big)+2F(u, v)\Big]dsd\xi \\ &+\frac{1}{2}\omega\int_0^t\int_{r}^\infty\int_{B(\xi)}e^{-\omega\eta}\Big[|u_\eta|^2+|v_\eta|^2+\widetilde{\rho}\Big(|\nabla u|^2\Big)+\widetilde{\rho}\Big(|\nabla v|^2\Big)+2F(u, v)\Big]dsd\xi d\eta \\&+\int_0^t\int_{r}^\infty\int_{B(\xi)}e^{-\omega\eta}\Big[|u_\eta|^{m+1}+|v_\eta|^{n+1}\Big]dsd\xi d\eta \\ &\leq \Big[-\mathcal{E}(r_0, t)\Big]e^{-\frac{\omega}{\sqrt{m_1}}(r-R_0)}. \end{align}$

(2.14)

We summarize the above result as the following theorem.

Theorem 2.1. Let $(u, v)$ be solution of the (1.1), (1.2), (2.1), (2.2) in $\Omega(R_0)$ , and $\rho$ satisfies condition A1. Then for fixed $t$ , $(u, v)$ either grows exponentially or decays exponentially. Specifically, either (2.11) holds or (2.14) holds.

3. The Phragmén-Lindelöf type alternative result under A2

If the function $\rho$ satisfies the condition A2, we recompute the inequality (2.6) and (2.7). Therefore

$\begin{align} \Big|\mathcal{E}_1(r, t)\Big|&\leq\Big[\int_0^t\int_{B(r)}e^{-\omega\eta}\rho^2\Big(|\nabla u|^2\Big)|\nabla u|^2ds d\eta\cdot\int_0^t\int_{B(r)}e^{-\omega\eta}u_\eta^2ds d\eta\Big]^\frac{1}{2}\\ &\leq K_1\sqrt{K_1}\Big[\int_0^t\int_{B(r)}e^{-\omega\eta}\rho^{-1}\Big(|\nabla u|^2\Big)ds d\eta\cdot\int_0^t\int_{B(r)}e^{-\omega\eta}u_\eta^2ds d\eta\Big]^\frac{1}{2}\\ &\leq\frac{K_1\sqrt{K_1K_2}}{2}\Big[\int_0^t\int_{B(r)}e^{-\omega\eta}\widetilde{\rho}\Big(|\nabla u|^2\Big)ds d\eta+\int_0^t\int_{B(r)}e^{-\omega\eta}u_\eta^2ds d\eta\Big], \end{align}$

(3.1)

and

$\begin{align} \Big|\mathcal{E}_2(r, t)\Big|&\leq\frac{K_1\sqrt{K_1K_2}}{2}\Big[\int_0^t\int_{B(r)}e^{-\omega\eta}\widetilde{\rho}\Big(|\nabla v|^2\Big)ds d\eta+\int_0^t\int_{B(r)}e^{-\omega\eta}v_\eta^2dsd\eta\Big]. \end{align}$

(3.2)

Inserting (3.1) and (3.2) into (2.3) and combining (2.5), we have

$\begin{align} \Big|\mathcal{E}(r, t)\Big|\leq \frac{K_1\sqrt{K_1K_2}}{\omega}\Big[\frac{\partial}{\partial r}\mathcal{E}(r, t)\Big]. \end{align}$

(3.3)

By following a similar method to that used in Section 2, we can obtain the Phragmén-Lindelöf type alternative result.

Theorem 3.1. Let $(u, v)$ be solution of the (1.1), (1.2), (2.1), (2.2) in $\Omega(R_0)$ , and $\rho$ satisfies condition A1. Then for fixed $t$ , $(u, v)$ either grows exponentially or decays exponentially.

Remark 3.1. Clearly, the rates of growth or decay obtained in Theorems 2.1 and 3.1 depend on $\omega$ . Because $\omega$ can be chosen large enough, the rates of growth or decay of the solutions can become large as we want.

Remark 3.2. The analysis in Sections 2 and 3 can be adapted to the single-wave equation

$\begin{align} u_{tt}-div\Big(\rho\Big(|\nabla u|^2\Big)\nabla u\Big)+|u_t|^{m-1}u_t+f(u) = 0 \end{align}$

(3.4)

and the heat conduction at low temperature

$\begin{align} au_{tt}+bu_t-c\Delta u+\Delta u_t = 0, \end{align}$

(3.5)

where $a, b, c > 0$ .

4. The Phragmén-Lindelöf type alternative result when $\rho(q^2)=b_1+b_2q^{2\beta}$

In this section, we suppose that $\rho$ satisfies $\rho(q^2) = b_1+b_2q^{2\beta}$ , where $b_1, b_2$ and $\beta$ are positive constants. Clearly, $\rho(q^2) = b_1+b_2q^{2\beta}$ can not satisfy A1 or A2. In this case, we define an "energy" function

$\begin{align} \mathcal{F}(r, t)& = \int_0^t\int_{B(r)}e^{-\omega\eta}\Big(b_1+b_2|\nabla u|^{2\beta}\Big)\nabla u\cdot\frac{\mathit{\boldsymbol{x}}}{r}u_\eta dsd\eta \\ &+\int_0^t\int_{B(r)}e^{-\omega\eta}\Big(b_1+b_2|\nabla v|^{2\beta}\Big)\nabla v\cdot\frac{\mathit{\boldsymbol{x}}}{r}v_\eta dsd\eta \\ &\doteq \mathcal{F}_1(r, t)+\mathcal{F}_2(r, t). \end{align}$

(4.1)

Computing as that in (2.4) and (2.5), we can get

$\begin{align} \mathcal{F}(r, t)& = \mathcal{F}(r_0, t)+\frac{1}{2}e^{-\omega t}\int_{r_0}^r\int_{B(\xi)}\Big[|u_t|^2+b_1|\nabla u|^2+\frac{1}{\beta+1}b_2|\nabla u|^{2(\beta+1)}\Big]dsd\xi\\ &+\frac{1}{2}e^{-\omega t}\int_{r_0}^r\int_{B(\xi)}\Big[|v_t|^2+b_1|\nabla v|^2+\frac{1}{\beta+1}b_2|\nabla v|^{2(\beta+1)}+F(u, v)\Big]dsd\xi\\ &+\frac{1}{2}\omega\int_0^t\int_{r_0}^r\int_{B(\xi)}e^{-\omega\eta}\Big[|u_\eta|^2+b_1|\nabla u|^2+\frac{1}{\beta+1}b_2|\nabla u|^{2(\beta+1)}\Big]ds d\xi d\eta\\ &+\frac{1}{2}\omega\int_0^t\int_{r_0}^r\int_{B(\xi)}e^{-\omega\eta}\Big[|v_\eta|^2+b_1|\nabla v|^2+\frac{1}{\beta+1}b_2|\nabla v|^{2(\beta+1)}+F(u, v)\Big]dsd\xi d\eta\\ &+\int_0^t\int_{r_0}^r\int_{B(\xi)}e^{-\omega\eta}|u_\eta|^{m+1}dsd\xi d\eta+\int_0^t\int_{B(r)}e^{-\omega\eta}|v_\eta|^{n+1}dsd\xi d\eta, \end{align}$

(4.2)

and

$\begin{align} \frac{\partial}{\partial r}\mathcal{F}(r, t)& = \frac{1}{2}e^{-\omega t}\int_{B(r)}\Big[|u_t|^2+b_1|\nabla u|^2+\frac{1}{\beta+1}b_2|\nabla u|^{2(\beta+1)}\Big]ds\\ &+\frac{1}{2}e^{-\omega t}\int_{B(r)}\Big[|v_t|^2+b_1|\nabla v|^2+\frac{1}{\beta+1}b_2|\nabla v|^{2(\beta+1)}+F(u, v)\Big]ds\\ &+\frac{1}{2}\omega\int_0^t\int_{B(r)}e^{-\omega\eta}\Big[|u_\eta|^2+b_1|\nabla u|^2+\frac{1}{\beta+1}b_2|\nabla u|^{2(\beta+1)}\Big]ds d\eta\\ &+\frac{1}{2}\omega\int_0^t\int_{B(r)}e^{-\omega\eta}\Big[|v_\eta|^2+b_1|\nabla v|^2+\frac{1}{\beta+1}b_2|\nabla v|^{2(\beta+1)}+F(u, v)\Big]ds d\eta\\ &+\int_0^t\int_{B(r)}e^{-\omega\eta}|u_\eta|^{m+1}dsd\eta+\int_0^t\int_{B(r)}e^{-\omega\eta}|v_\eta|^{n+1}ds d\eta. \end{align}$

(4.3)

Using the Hölder inequality and Young's inequality, we have

$\begin{align} \Big|\mathcal{F}_1(r, t)\Big|&\leq b_1\Big[\int_0^t\int_{B(r)}e^{-\omega\eta}|\nabla u|^2dsd\eta \int_0^t\int_{B(r)}e^{-\omega\eta}u_\eta^2 dsd\eta\Big]^\frac{1}{2} \\ &+b_2\Big[\int_0^t\int_{B(r)}e^{-\omega\eta}|\nabla u|^{2(\beta+1)}dsd\eta\Big]^\frac{2\beta+1}{2(\beta+1)} \int_0^t\int_{B(r)}e^{-\omega\eta}|u_\eta|^{m+1} dsd\eta\Big]^\frac{1}{m+1} |t2\pi r|^{\frac{1}{2(\beta+1)}-\frac{1}{m+1}} \\ &\leq \frac{\sqrt{b_1}}{R_0\omega}\Big[\frac{b_1}{2}r\omega\int_0^t\int_{B(r)}e^{-\omega\eta}|\nabla u|^2dsd\eta+ \frac{1}{2}r\omega\int_0^t\int_{B(r)}e^{-\omega\eta}u_\eta^2 dsd\eta\Big] \\ &+b_2|2t\pi|^{\frac{1}{2(\beta+1)}-\frac{1}{m+1}}R_0^{-\frac{\beta}{\beta+1}-\frac{2}{m+1}}\Big[\frac{\frac{2\beta+1}{2(\beta+1)}r}{\frac{2\beta+1}{2(\beta+1)} +\frac{1}{m+1}}\int_0^t\int_{B(r)}e^{-\omega\eta}|\nabla u|^{2(\beta+1)}dsd\eta \\ &+\frac{\frac{1}{m+1}r}{\frac{2\beta+1}{2(\beta+1)} +\frac{1}{m+1}} \int_0^t\int_{B(r)}e^{-\omega\eta}|u_\eta|^{m+1} dsd\eta\Big]^{\frac{2\beta+1}{2(\beta+1)} +\frac{1}{m+1}}, \end{align}$

(4.4)

and

$\begin{align} \Big|\mathcal{F}_2(r, t)\Big|&\leq\frac{\sqrt{b_1}}{R_0\omega}\Big[\frac{b_1}{2}r\omega\int_0^t\int_{B(r)}e^{-\omega\eta}|\nabla v|^2dsd\eta+ \frac{1}{2}r\omega\int_0^t\int_{B(r)}e^{-\omega\eta}v_\eta^2 dsd\eta\Big] \\ &+b_2|2t\pi|^{\frac{1}{2(\beta+1)}-\frac{1}{n+1}}R_0^{-\frac{\beta}{\beta+1}-\frac{2}{n+1}} \Big[\frac{\frac{2\beta+1}{2(\beta+1)}r}{\frac{2\beta+1}{2(\beta+1)} +\frac{1}{n+1}}\int_0^t\int_{B(r)}e^{-\omega\eta}|\nabla v|^{2(\beta+1)}dsd\eta \\ &+\frac{\frac{1}{n+1}r}{\frac{2\beta+1}{2(\beta+1)} +\frac{1}{n+1}} \int_0^t\int_{B(r)}e^{-\omega\eta}|v_\eta|^{n+1} dsd\eta\Big]^{\frac{2\beta+1}{2(\beta+1)} +\frac{1}{n+1}}, \end{align}$

(4.5)

where we have chosen $\frac{1}{2(\beta+1)} > \frac{1}{n+1}$ . Inserting (4.4) and (4.5) into (4.1), we obtain

$\begin{align} \Big|\mathcal{F}(r, t)\Big|&\leq c_1\Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big]+c_2\Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big]^{\frac{2\beta+1}{2(\beta+1)} +\frac{1}{m+1}} +c_3\Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big]^{\frac{2\beta+1}{2(\beta+1)} +\frac{1}{n+1}}, \end{align}$

(4.6)

where $c_1 = \frac{\sqrt{b_1}}{R_0\omega}, c_2 = b_2|2t\pi|^{\frac{1}{2(\beta+1)}-\frac{1}{n+1}}R_0^{-\frac{\beta}{\beta+1}-\frac{2}{m+1}}, c_3 = b_2|2t\pi|^{\frac{1}{2(\beta+1)}-\frac{1}{n+1}}R_0^{-\frac{\beta}{\beta+1}-\frac{2}{n+1}}$ .

Next, we will analyze Eq (4.6) in two cases

I. If $\exists r_0\geq R_0$ such that $\mathcal{F}(r_0, t)\geq0$ , then $\mathcal{F}(r, t)\geq\mathcal{F}(r_0, t)\geq0, r\geq r_0.$ Therefore, (4.6) can be rewritten as

$\begin{align} \mathcal{F}(r, t)&\leq c_1\Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big]+c_2\Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big]^{\frac{2\beta+1}{2(\beta+1)} +\frac{1}{m+1}} \\ &+c_3\Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big]^{\frac{2\beta+1}{2(\beta+1)} +\frac{1}{n+1}}, \quad\quad\quad\quad\quad r\geq r_0. \end{align}$

(4.7)

Using Young's inequality, we have

$\begin{align} \Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big]^{\frac{2\beta+1}{2(\beta+1)} +\frac{1}{m+1}}&\leq\Big(\frac{4\beta+3}{4(\beta+1)}+\frac{1}{2(m+1)}\Big) \Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big]^\frac{1}{2} \\ &+\Big(\frac{1}{4(\beta+1)}-\frac{1}{2(m+1)}\Big) \Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big], \end{align}$

(4.8)

$\begin{align} \Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big]^{\frac{2\beta+1}{2(\beta+1)} +\frac{1}{n+1}}&\leq\Big(\frac{4\beta+3}{4(\beta+1)}+\frac{1}{2(n+1)}\Big) \Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big]^\frac{1}{2} \\ &+\Big(\frac{1}{4(\beta+1)}-\frac{1}{2(n+1)}\Big) \Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big]. \end{align}$

(4.9)

Inserting (4.8) and (4.9) into (4.7), we have

$\begin{align} \mathcal{F}(r, t)&\leq c_4\Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big]^\frac{1}{2}+c_5\Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big], \quad r\geq r_0. \end{align}$

(4.10)

where $c_4 = c_2\Big(\frac{4\beta+3}{4(\beta+1)}+\frac{1}{2(m+1)}\Big)+c_3\Big(\frac{4\beta+3}{4(\beta+1)}+\frac{1}{2(n+1)}\Big)$ and $c_5 = c_1+c_2\Big(\frac{1}{4(\beta+1)}-\frac{1}{2(m+1)}\Big)+c_3\Big(\frac{1}{4(\beta+1)}-\frac{1}{2(n+1)}\Big)$ . From (4.10) we have

$\begin{align} \frac{\mathcal{F}(r, t)}{c_5}&\leq \Big[\sqrt{r\frac{\partial}{\partial r}\mathcal{F}(r, t)}+\frac{c_4}{2c_5}\Big]^2 -\frac{c^2_4}{4c^2_5}, \quad r\geq r_0 \end{align}$

$\begin{align} \frac{\frac{\partial}{\partial r}\mathcal{F}(r, t)}{\Big[\sqrt{\frac{\mathcal{F}(r, t)}{c_5}+\frac{c^2_4}{4c^2_5}}-\frac{c_4}{2c_5}\Big]^2} \geq \frac{1}{r}, \quad r\geq r_0 \end{align}$

(4.11)

Integrating (4.11) from $r_0$ to $r$ , we get

$\begin{align} 2c_5\Big[ln\Big(\sqrt{\frac{\mathcal{F}(r, t)}{c_5}+\frac{c^2_4}{4c^2_5}}-\frac{c_4}{2c_5}\Big) &-ln\Big(\sqrt{\frac{\mathcal{F}(r_0, t)}{c_5}+\frac{c^2_4}{4c^2_5}}-\frac{c_4}{2c_5}\Big)\Big] \\ -c_4\Big\{\Big[\sqrt{\frac{\mathcal{F}(r, t)}{c_5}+\frac{c^2_4}{4c^2_5}}-\frac{c_4}{2c_5}\Big]^{-1} &-\Big[\sqrt{\frac{\mathcal{F}(r_0, t)}{c_5}+\frac{c^2_4}{4c^2_5}}-\frac{c_4}{2c_5}\Big]^{-1}\Big\} \\ &\geq ln\Big(\frac{r}{r_0}\Big), \quad r\geq r_0 \end{align}$

(4.12)

Dropping the third term on the left of (4.12), we have

$\begin{align} \sqrt{\frac{\mathcal{F}(r, t)}{c_5}+\frac{c^2_4}{4c^2_5}}-\frac{c_4}{2c_5}\geq Q_1(r_0, t)\Big(\frac{r}{r_0}\Big)^\frac{1}{2c_5}, \end{align}$

(4.13)

where $Q_1(r_0, t) = \Big(\sqrt{\frac{\mathcal{F}(r_0, t)}{c_5}+\frac{c^2_4}{4c^2_5}}-\frac{c_4}{2c_5}\Big)\exp\Big\{-\frac{c_4}{2c_5} \Big[\sqrt{\frac{\mathcal{F}(r_0, t)}{c_5}+\frac{c^2_4}{4c^2_5}}-\frac{c_4}{2c_5}\Big]^{-1}\Big\}$ .

In view of

$\sqrt{\frac{\mathcal{F}(r, t)}{c_5}+\frac{c^2_4}{4c^2_5}}\leq\sqrt{\frac{\mathcal{F}(r, t)}{c_5}}+\frac{c_4}{2c_5},$

we have from (4.13)

$\begin{align} \mathcal{F}(r, t)\geq c_5Q_1^2(r_0, t)\Big(\frac{r}{r_0}\Big)^\frac{1}{c_5}. \end{align}$

(4.14)

Combining (4.2) and (4.14), we have

$\begin{align} \lim\limits_{r\rightarrow \infty}&\Big\{r^{-\frac{1}{c_5}}\Big[\frac{1}{2}e^{-\omega t}\int_{r_0}^r\int_{B(\xi)}\Big[|u_t|^2+b_1|\nabla u|^2+\frac{1}{\beta+1}b_2|\nabla u|^{2(\beta+1)}\Big]dsd\xi\\ &+\frac{1}{2}e^{-\omega t}\int_{r_0}^r\int_{B(\xi)}\Big[|v_t|^2+b_1|\nabla v|^2+\frac{1}{\beta+1}b_2|\nabla v|^{2(\beta+1)}+F(u, v)\Big]dsd\xi \\ &+\frac{1}{2}\omega\int_0^t\int_{r_0}^r\int_{B(\xi)}e^{-\omega\eta}\Big[|u_\eta|^2+b_1|\nabla u|^2+\frac{1}{\beta+1}b_2|\nabla u|^{2(\beta+1)}\Big]ds d\xi d\eta\\ &+\frac{1}{2}\omega\int_0^t\int_{r_0}^r\int_{B(\xi)}e^{-\omega\eta}\Big[|v_\eta|^2+b_1|\nabla v|^2+\frac{1}{\beta+1}b_2|\nabla v|^{2(\beta+1)}+F(u, v)\Big]dsd\xi d\eta \\ &+\int_0^t\int_{r_0}^r\int_{B(\xi)}e^{-\omega\eta}|u_\eta|^{m+1}dsd\xi d\eta+\int_0^t\int_{B(r)}e^{-\omega\eta}|v_\eta|^{n+1}dsd\xi d\eta\Big]\Big\} \\ &\geq c_5Q_1^2(r_0, t)r_0^{-\frac{1}{c_5}}. \end{align}$

(4.15)

II. If $\forall r\geq R_0$ such that $\mathcal{F}(r, t) < 0$ , then (4.6) can be rewritten as

$\begin{align} -\mathcal{F}(r, t)&\leq c_1\Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big]+c_2\Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big]^{\frac{2\beta+1}{2(\beta+1)} +\frac{1}{m+1}} \\ &+c_3\Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big]^{\frac{2\beta+1}{2(\beta+1)} +\frac{1}{n+1}}, \quad\quad\quad\quad\quad r\geq R_0. \end{align}$

(4.16)

Without losing generality, we suppose that $m > n > 1$ .

$\begin{align} \Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big]^{\frac{2\beta+1}{2(\beta+1)} +\frac{1}{n+1}}&\leq\frac{\frac{1}{n+1}-\frac{1}{m+1}}{\frac{2\beta+1}{2(\beta+1)}-\frac{1}{m+1}} \Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big] \\ &+\frac{\frac{2\beta+1}{2(\beta+1)}-\frac{1}{m+1}}{\frac{2\beta+1}{2(\beta+1)}-\frac{1}{m+1}} \Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big]^{\frac{2\beta+1}{2(\beta+1)} +\frac{1}{m+1}}, \end{align}$

(4.17)

$\begin{align} \Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big]&\leq\frac{2\beta(m+3)+4}{(2\beta+1)(m+2)+1} \Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big]^{\frac{2\beta+1}{2(\beta+1)} +\frac{1}{m+1}} \\ &+\frac{m+1-2(\beta+1)}{(2\beta+1)(m+2)+1} \Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big]^{2\Big[\frac{2\beta+1}{2(\beta+1)} +\frac{1}{m+1}\Big]}. \end{align}$

(4.18)

Inserting (4.17) and (4.18) into (4.16), we get

$\begin{align} -\mathcal{F}(r, t)&\leq c_6\Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big]^{\frac{2\beta+1}{2(\beta+1)} +\frac{1}{m+1}} +c_7\Big[r\frac{\partial}{\partial r}\mathcal{F}(r, t)\Big]^{2\Big[\frac{2\beta+1}{2(\beta+1)} +\frac{1}{m+1}\Big]}, \ r\geq R_0, \end{align}$

(4.19)

where $c_6 = \Big[c_1+c_3\frac{\frac{1}{n+1}-\frac{1}{m+1}}{\frac{2\beta+1}{2(\beta+1)}-\frac{1}{m+1}}\Big] \frac{2\beta(m+3)+4}{(2\beta+1)(m+2)+1}+c_3\frac{\frac{2\beta+1}{2(\beta+1)}-\frac{1}{m+1}}{\frac{2\beta+1}{2(\beta+1)}-\frac{1}{m+1}}, c_7 = c_3\frac{m+1-2(\beta+1)}{(2\beta+1)(m+2)+1}$ . From (4.19) we obtain

$\begin{align} r\frac{\partial}{\partial r}\mathcal{F}(r, t)&\geq \Big[\sqrt{\frac{-\mathcal{F}(r, t)}{c_7}+\frac{c_6^2}{4c_7^2}}-\frac{c_6}{2c_7}\Big]^{\frac{1}{\frac{2\beta+1}{2(\beta+1)} +\frac{1}{m+1}}}, \ r\geq R_0 \end{align}$

$\begin{align} 2c_7\frac{\sqrt{\frac{-\mathcal{F}(r, t)}{c_7}+\frac{c_6^2}{4c_7^2}}}{\Big[\sqrt{\frac{-\mathcal{F}(r, t)}{c_7}+\frac{c_6^2}{4c_7^2}}-\frac{c_6}{2c_7}\Big]^{\frac{1}{\frac{2\beta+1}{2(\beta+1)} +\frac{1}{m+1}}}}d\Big\{\sqrt{\frac{-\mathcal{F}(r, t)}{c_7}+\frac{c_6^2}{4c_7^2}}-\frac{c_6}{2c_7}\Big\}&\leq-\frac{1}{r}, \ r\geq R_0 \end{align}$

(4.20)

Integrating (4.20) from $R_0$ to $r$ , we obtain

$\begin{align} &\frac{2c_7\Big(\frac{2\beta+1}{2(\beta+1)}+\frac{1}{m+1}\Big)}{\frac{\beta}{2(\beta+1)}+\frac{2}{m+1}} \Big[\sqrt{\frac{-\mathcal{F}(r, t)}{c_7}+\frac{c_6^2}{4c_7^2}}-\frac{c_6}{2c_7}\Big]^\frac{\frac{\beta}{2(\beta+1)}+\frac{2}{m+1}}{\frac{2\beta+1}{2(\beta+1)}+\frac{1}{m+1}} \\ &-\frac{2c_7\Big(\frac{2\beta+1}{2(\beta+1)}+\frac{1}{m+1}\Big)}{\frac{\beta}{2(\beta+1)}+\frac{2}{m+1}} \Big[\sqrt{\frac{-\mathcal{F}(R_0, t)}{c_7}+\frac{c_6^2}{4c_7^2}}-\frac{c_6}{2c_7}\Big]^\frac{\frac{\beta}{2(\beta+1)}+\frac{2}{m+1}}{\frac{2\beta+1}{2(\beta+1)}+\frac{1}{m+1}} \\ &-\frac{c_6\Big(\frac{2\beta+1}{2(\beta+1)}+\frac{1}{m+1}\Big)}{\frac{1}{2(\beta+1)}-\frac{1}{m+1}} \Big[\sqrt{\frac{-\mathcal{F}(r, t)}{c_7}+\frac{c_6^2}{4c_7^2}}-\frac{c_6}{2c_7}\Big]^{-\frac{\frac{1}{2(\beta+1)}-\frac{1}{m+1}}{\frac{2\beta+1}{2(\beta+1)}+\frac{1}{m+1}}} \\ &+\frac{c_6\Big(\frac{2\beta+1}{2(\beta+1)}+\frac{1}{m+1}\Big)}{\frac{1}{2(\beta+1)}-\frac{1}{m+1}} \Big[\sqrt{\frac{-\mathcal{F}(R_0, t)}{c_7}+\frac{c_6^2}{4c_7^2}}-\frac{c_6}{2c_7}\Big]^{-\frac{\frac{1}{2(\beta+1)}-\frac{1}{m+1}}{\frac{2\beta+1}{2(\beta+1)}+\frac{1}{m+1}}} \\ &\leq ln\Big(\frac{R_0}{r}\Big). \end{align}$

(4.21)

Dropping the first and fourth terms on the left of (4.21), we obtain

$\begin{align} \sqrt{\frac{-\mathcal{F}(r, t)}{c_7}+\frac{c_6^2}{4c_7^2}} \leq\Big[c_8ln\Big(\frac{r}{R_0}\Big) -Q_2(R_0, t)\Big]^{-\frac{\frac{2\beta+1}{2(\beta+1)}+\frac{1}{m+1}}{\frac{1}{2(\beta+1)}-\frac{1}{m+1}}} +\frac{c_6}{2c_7}, \end{align}$

(4.22)

where $Q_2(R_0, t) = \frac{2c_7\Big(\frac{1}{2(\beta+1)}-\frac{1}{m+1}\Big)}{c_6\Big(\frac{\beta}{2(\beta+1)}+\frac{2}{m+1}\Big)} \Big[\sqrt{\frac{-\mathcal{F}(R_0, t)}{c_7}+\frac{c_6^2}{4c_7^2}}-\frac{c_6}{2c_7}\Big]^\frac{\frac{\beta}{2(\beta+1)}+\frac{2}{m+1}}{\frac{2\beta+1}{2(\beta+1)}+\frac{1}{m+1}}$ and $c_8 = \frac{\frac{1}{2(\beta+1)}-\frac{1}{m+1}}{c_6\Big(\frac{2\beta+1}{2(\beta+1)} +\frac{1}{m+1}\Big)}$ .

Squaring (4.22) we have

$\begin{align} -\mathcal{F}(r, t)& \leq c_7\Big[c_8ln\Big(\frac{r}{R_0}\Big) -Q_2(R_0, t)\Big]^{-\frac{\frac{2\beta+1}{\beta+1}+\frac{2}{m+1}}{\frac{1}{2(\beta+1)}-\frac{1}{m+1}}} \\ &+c_6\Big[c_8ln\Big(\frac{r}{R_0}\Big) -Q_2(R_0, t)\Big]^{-\frac{\frac{2\beta+1}{2(\beta+1)}+\frac{1}{m+1}}{\frac{1}{2(\beta+1)}-\frac{1}{m+1}}}. \end{align}$

(4.23)

From (4.15) and (4.23) we can obtain the following theorem.

Theorem 4.1. Let $(u, v)$ be the solution of (1.1), (1.2), (2.1) and (2.2) with $\rho(q^2) = b_1+b_2q^{2\beta}$ , where $\frac{1}{2(\beta+1)} > \max\Big\{\frac{1}{m+1}, \frac{1}{n+1}\Big\}$ . Then for fixed $t$ , when $r\rightarrow \infty$ , $(u, v)$ either grows algebraically or decays logarithmically. The growth rate is at least as fast as $z^\frac{1}{c_5}$ and the decay rate is at least as fast as $\Big(ln r\Big)^{-\frac{\frac{2\beta+1}{2(\beta+1)}+\frac{1}{m+1}}{\frac{1}{2(\beta+1)}-\frac{1}{m+1}}}$ .

Remark 4.1. Obviously, in this case of $\rho(q^2) = b_1+b_2q^{2\beta}$ , the decay rate obtained by Theorem 4.1 is slower than that obtained by Theorem 2.1 and Theorem 3.1.

5. A nonlinear system of viscoelastic type

In this section, we concern with a system of two coupled viscoelastic equations

$\begin{align} u_{tt}&-\Delta u+\int_0^th_1(t-\eta)\Delta u(\eta)d\eta+f_1(u, v) = 0, \end{align}$

(5.1)

$\begin{align} v_{tt}&-\Delta v+\int_0^th_2(t-\eta)\Delta v(\eta)d\eta+f_2(u, v) = 0, \end{align}$

(5.2)

which describes the interaction between two different fields arising in viscoelasticity. In (5.1) and (5.2), $0 < t < T$ and $h_1, h_2$ are differentiable functions satisfying $h_1(0), h_2(0) > 0$ and

$\begin{align} 2\Big(&\int_0^T h_1^2(T-\eta)d\eta\Big), \ \frac{4T}{h_1(0)}\Big[\int_0^T \Big(h_1'(T-\eta)-\omega h_1(T-\eta)\Big)^2d\eta\Big] \leq h_1(0), \end{align}$

(5.3)

$\begin{align} 2\Big(&\int_0^T h_2^2(T-\eta)d\eta\Big), \ \frac{4T}{h_2(0)}\Big[\int_0^T \Big(h_2'(T-\eta)-\omega h_2(T-\eta)\Big)^2d\eta\Big] \leq h_2(0). \end{align}$

(5.4)

Messaoudi and Tatar ^[24] considered the system (5.1) and (5.2) in a bounded domain and proved the uniform decay for the solution when $t\rightarrow \infty$ . For more special cases, one can refer to ^[25,26,27]. They mainly concerned the well-posedness of the solutions and proved that the solutions decayed uniformly under some suitable conditions. However, the present paper extends the previous results to Eqs (5.1) and (5.2) in an exterior region. We consider Eqs (5.1) and (5.2) with the initial-boundary conditions (2.1) and (2.2) in $\Omega$ .

We define two functions

$\begin{align} G_1(r, t)& = \int_0^t\int_{B(r)}e^{-\omega\eta}\nabla u\cdot\frac{\mathit{\boldsymbol{x}}}{r}u_\eta dsd\eta -\int_0^t\int_{B(r)}e^{-\omega\eta}\Big(\int_0^\eta h_1(\eta-s)\nabla uds\Big)\cdot\frac{\mathit{\boldsymbol{x}}}{r}u_\eta dsd\eta \\ &\doteq I_1+I_2, \end{align}$

(5.5)

$\begin{align} G_2(r, t)& = \int_0^t\int_{B(r)}e^{-\omega\eta}\nabla v\cdot\frac{\mathit{\boldsymbol{x}}}{r}v_\eta dsd\eta -\int_0^t\int_{B(r)}e^{-\omega\eta}\Big(\int_0^\eta h_2(\eta-s)\nabla vds\Big)\cdot\frac{\mathit{\boldsymbol{x}}}{r}v_\eta dsd\eta \\ &\doteq J_1+J_2. \end{align}$

(5.6)

Integrating (5.5) from $r_0$ to $r$ and using (5.1), (5.2), (2.1), (2.2) and the divergence theorem, we have

$\begin{align} G_1(r, t)& = G_1(r_0, t)+\frac{1}{2}\int_{r_0}^r\int_{B(\xi)}e^{-\omega t}\Big[|u_t|^2+|\nabla u|^2\Big]dsd\xi+\frac{1}{2}\omega\int_0^t\int_{r_0}^r\int_{B(\xi)}e^{-\omega \eta}|u_\eta|^2dsd\xi d\eta\\ &+\frac{1}{2}\omega\int_0^t\int_{r_0}^r\int_{B(\xi)}e^{-\omega \eta}|\nabla u|^2dsd\xi d\eta+\int_0^t\int_{r_0}^r\int_{B(\xi)}e^{-\omega \eta}h_1(0)|\nabla u|^2dsd\xi d\eta\\ &-\int_{r_0}^r\int_{B(\xi)}e^{-\omega t}\Big(\int_0^t h_1(t-\tau)\nabla ud\tau\Big)\cdot\nabla udsd\xi\\ &+\int_0^t\int_{r_0}^r\int_{B(\xi)}e^{-\omega \eta}\Big[\int_0^\eta \Big(h_1'(\eta-\tau)-\omega h_1(\eta-\tau)\Big)\nabla ud\tau\Big]\cdot\nabla udsd\xi d\eta\\ &+\int_0^t\int_{r_0}^r\int_{B(\xi)}e^{-\omega \eta}f_1(u, v)u_\eta dsd\xi d\eta. \end{align}$

(5.7)

From (5.7) it follows that

$\begin{align} \frac{\partial}{\partial z}G_1(r, t)& = \frac{1}{2}\int_{B(r)}e^{-\omega t}\Big[|u_t|^2+|\nabla u|^2\Big]ds+\frac{1}{2}\omega\int_0^t\int_{B(r)}e^{-\omega \eta}|u_\eta|^2dsd\eta \\ &+\frac{1}{2}\omega\int_0^t\int_{B(r)}e^{-\omega \eta}|\nabla u|^2dsd\eta+\int_0^t\int_{B(r)}e^{-\omega \eta}h_1(0)|\nabla u|^2dsd\eta \\ &-\int_{B(r)}e^{-\omega t}\Big(\int_0^t h_1(t-\tau)\nabla ud\tau\Big)\cdot\nabla uds\\ &+\int_0^t\int_{B(r)}e^{-\omega \eta}\Big[\int_0^\eta \Big(h_1'(\eta-\tau)-\omega h_1(\eta-\tau)\Big)\nabla ud\tau\Big]\cdot\nabla uds d\eta\\ &+\int_0^t\int_{B(r)}e^{-\omega \eta}f_1(u, v)u_\eta dsd\eta. \end{align}$

(5.8)

By the Young inequality and the Hölder inequality, we have

$\begin{align} \Big|&-\int_{B(r)}e^{-\omega t}\Big(\int_0^t h_1(\eta-\tau)\nabla ud\tau\Big)\cdot\nabla uds\Big|\\ &\leq\int_{B(r)}e^{-\omega t}\Big[\Big(\int_0^t h_1(t-s)\nabla u(s)ds\Big)^2+\frac{1}{4}|\nabla u|^2\Big]ds\\ &\leq\Big(\int_0^t h_1^2(t-\tau)d\tau\Big)\int_0^t\int_{B(r)}e^{-\omega \eta}|\nabla u|^2dsd\eta+\frac{1}{4}\int_{B(r)}e^{-\omega t}|\nabla u|^2ds, \end{align}$

(5.9)

and

$\begin{align} \Big|&\int_0^t\int_{B(r)}e^{-\omega \eta}\Big[\int_0^\eta \Big(h_1'(\eta-\tau)-\omega h_1(\eta-\tau)\Big)\nabla ud\tau\Big]\cdot\nabla uds d\eta\Big|\\ &\leq \frac{t}{h_1(0)}\Big[\int_0^t \Big(h_1'(\eta-\tau)-\omega h_1(\eta-\tau)\Big)^2d\tau\Big]\int_0^t\int_{B(r)}e^{-\omega \eta}|\nabla u|^2dsd\eta\\ &+\frac{1}{4}\int_0^t\int_{B(r)}e^{-\omega \eta}h_1(0)|\nabla u|^2dsd\eta. \end{align}$

(5.10)

Inserting (5.9) and (5.10) into (5.8) and using (5.3), we have

$\begin{align} \frac{\partial}{\partial z}G_1(r, t)&\geq\frac{1}{2}\int_{B(r)}e^{-\omega t}\Big[|u_t|^2+\frac{1}{2}|\nabla u|^2\Big]ds+\frac{1}{2}\omega\int_0^t\int_{B(r)}e^{-\omega \eta}|u_\eta|^2ds d\eta\\ &+\frac{1}{2}\omega\int_0^t\int_{B(r)}e^{-\omega \eta}|\nabla u|^2dsd\eta+\int_0^t\int_{B(r)}e^{-\omega \eta}f_1(u, v)u_\eta ds d\eta. \end{align}$

(5.11)

Similar to (5.11), we also have for $G_2(r, t)$

$\begin{align} \frac{\partial}{\partial z}G_2(r, t)&\geq\frac{1}{2}\int_{B(r)}e^{-\omega t}\Big[|v_t|^2+\frac{1}{2}|\nabla v|^2\Big]ds+\frac{1}{2}\omega\int_0^t\int_{B(r)}e^{-\omega \eta}|v_\eta|^2dsd\eta\\ &+\frac{1}{2}\omega\int_0^t\int_{B(r)}e^{-\omega \eta}|\nabla v|^2dsd\eta+\int_0^t\int_{B(r)}e^{-\omega \eta}f_2(u, v)v_\eta dsd\eta. \end{align}$

(5.12)

If we define

$G(r, t) = G_1(r, t)+G_2(r, t),$

then by (5.11) and (5.12) we have

$\begin{align} \frac{\partial}{\partial z}G(r, t)&\geq\frac{1}{2}\int_{B(r)}e^{-\omega t}\Big[|u_t|^2+|v_t|^2+\frac{1}{2}|\nabla u|^2+\frac{1}{2}|\nabla v|^2+2F(u, v)\Big]ds \\ &+\frac{1}{2}\omega\int_0^t\int_{B(r)}e^{-\omega \eta}\Big[|u_\eta|^2+|v_\eta|^2+|\nabla u|^2+|\nabla v|^2+2F(u, v)\Big]dsd\eta. \end{align}$

(5.13)

On the other hand, we bound $G(r, t)$ by $\frac{\partial}{\partial r}G(r, t)$ . Using the Hölder inequality, the AG mean inequality, (5.3) and combining (5.13), we have

$\begin{align} \Big|I_1\Big|+\Big|J_1\Big|&\leq\Big(\int_0^t\int_{B(r)}e^{-\omega\eta}|\nabla u|^2 dAd\eta\cdot\int_0^t\int_{B(r)}e^{-\omega\eta}|u_\eta|^2 dsd\eta\Big)^\frac{1}{2} \\ &+\Big(\int_0^t\int_{B(r)}e^{-\omega\eta}|\nabla v|^2 dAd\eta\cdot\int_0^t\int_{B(r)}e^{-\omega\eta}|v_\eta|^2 dsd\eta\Big)^\frac{1}{2} \\ &\leq\frac{1}{\omega}\Big[\frac{\partial}{\partial r}G(r, t)\Big], \end{align}$

(5.14)

and

$\begin{align} \Big|I_2\Big|&\leq\Big(\int_0^t\int_{B(r)}e^{-\omega\eta}\Big|\int_0^\eta h_1(\eta-\tau)\nabla ud\tau\Big|^2 dsd\eta\cdot\int_0^t\int_{B(r)}e^{-\omega\eta}|u_\eta|^2dsd\eta\Big)^\frac{1}{2} \\ &\leq\Big(\int_0^t\int_{B(r)}e^{-\omega\eta}\Big(\int_0^\eta h^2_1(\eta-\tau)d\tau\Big)\Big(\int_0^\eta |\nabla u|^2d\tau\Big) dsd\eta\cdot\int_0^t\int_{B(r)}e^{-\omega\eta}|u_\eta|^2dsd\eta\Big)^\frac{1}{2} \\ &\leq t\Big(\int_0^t h^2_1(\eta-\tau)d\tau\Big)^\frac{1}{2}\Big(\int_0^t\int_{B(r)}e^{-\omega\eta} |\nabla u|^2 dsd\eta\cdot\int_0^t\int_{B(r)}e^{-\omega\eta}|u_\eta|^2dsd\eta\Big)^\frac{1}{2} \\ &\leq\frac{t}{2}\Big(\int_0^t h^2_1(\eta-\tau)d\tau\Big)^\frac{1}{2}\Big[\int_0^t\int_{B(r)}e^{-\omega\eta} |\nabla u|^2 dsd\eta+\int_0^t\int_{B(r)}e^{-\omega\eta}|u_\eta|^2dsd\eta\Big]\\ &\leq\frac{T}{2\omega}\sqrt{h_1(0)}\Big[\frac{\partial}{\partial r}G(r, t)\Big]. \end{align}$

(5.15)

Similar to (5.15), we have

$\begin{align} \Big|J_2\Big|\leq\frac{T}{2\omega}\sqrt{h_2(0)}\Big[\frac{\partial}{\partial r}G(r, t)\Big]. \end{align}$

(5.16)

Inserting (5.14)–(5.16) into (5.5) and (5.6), we have

$\begin{align} \Big|G(r, t)\Big|\leq c_9\frac{1}{\omega}\Big[\frac{\partial}{\partial r}G(r, t)\Big], \end{align}$

(5.17)

where $c_9 = \frac{T}{2}(\sqrt{h_1(0)}+\sqrt{h_2(0)})+2$ .

We can follow the similar arguments given in the previous sections to obtain the following theorem.

Theorem 5.1. Let $(u, v)$ be the solution of (5.1), (5.2), (2.1) and (2.2) in $\Omega$ , and (5.3) and (5.4) hold. For fixed $t$ ,

(1) If $\exists\ R_0\geq r_0$ , $G(R_0, t)\geq0$ , then

$G(r, t)\geq G(R_0, t)e^{\frac{\omega}{b_9}(r-R_0)}.$

(2) If $\forall\ r\geq r_0$ , $G(r, t) < 0$ , then

$-G(r, t)\leq \Big[-G(r_0, t)\Big]e^{-\frac{\omega}{b_9}(r-r_0)}.$

Again, the rate of growth or decay obtained in this case is arbitrarily large

Remark 5.1. It is clear that the above analysis can be adapted without difficulties to the equation (see ^[28,29])

$u_{tt}-k_0\triangle u+\int_0^tdiv[a(x)h(t-s)\nabla u(s)]ds+b(x)g(u_t)+f(u) = 0$

and the equation (see ^[30])

$|u_{t}|^\sigma u_{tt}-k_0\triangle u-\Delta u_{tt}+\int_0^th(t-s)\Delta u(s)]ds-\gamma\Delta u_t = 0$

with some suitable $g$ and $a(x)+b(x)\geq b_{10} > 0$ and $k_0, \sigma, \gamma > 0.$

6. Conclusions

In this paper, we have considered several situations where the solutions of Eqs (1.1) and (1.2) either grow or decay exponentially or polynomially. We emphasize that the Poincaré inequality on the cross sections is not used in this paper. Thus, our results also hold for the two-dimensional case. On the other hand, there are some deeper problems to be studied in this paper. We can continue to study the continuous dependence of coefficients in the equation as that in ^[31]. These are the issues we will continue to study in the future.

Use of AI tools declaration

The authors declare they have not used Artificial Intelligence (AI) tools in the creation of this article.

Acknowledgments

The authors express their heartfelt thanks to the editors and referees who have provided some important suggestions. This work was supported by the Guangzhou Huashang College Tutorial Scientific Research Project (2022HSDS09), National Natural Science Foundation of China (11371175) and the Research Team Project of Guangzhou Huashang College (2021HSKT01).

Conflict of interest

The authors declare that there is no conflict of interest.

References

[1]	Regulatory framework. Moroccan Agency For Energy Efficiency (AMEE). Available from: https://www.amee.ma/fr/cadre-reglementaire.
[2]	Lairgi L, Daya A, Elotmani R, et al. (2019) Contribution to the study of the energy needs for a building in the city of errachidia influence of orientation and local materials. Key Eng Mater 820: 18–28. https://doi.org/10.4028/www.scientific.net/KEM.820.18 doi: 10.4028/www.scientific.net/KEM.820.18
[3]	Idchabani R, Khyad A, El Ganaoui M (2017) Optimizing insulation thickness of external walls in cold region of Morocco based on life cycle cost analysis. Energy Procedia 139: 117–121. https://doi.org/10.1016/J.EGYPRO.2017.11.183 doi: 10.1016/j.egypro.2017.11.183
[4]	Louanate A, Otmani RE, Kandoussi K, et al. (2021) Energy saving potential of phase change materials-enhanced building envelope considering the six Moroccan climate zones. J Build Phys 45: 482–506. https://doi.org/10.1177/17442591211006444 doi: 10.1177/17442591211006444
[5]	Khan MY, Baqi A, Talib A (2020) Energy efficiency analysis of a building envelope. In: Bose, M., Modi, A., Proceedings of the 7th International Conference on Advances in Energy Research, Singapore: Springer, 1691–1702. https://doi.org/10.1007/978-981-15-5955-6_160
[6]	Lairgi L, Lagtayi R, Daya A, et al. (2021) The impact of arid climate on the indoor thermal comfort in the south-east of Morocco. Int J Photoenergy 2021. https://doi.org/10.1155/2021/5554629 doi: 10.1155/2021/5554629
[7]	Lagtayi R, Lairgi L, Khouya A, et al. (2021) Acoustic and thermal correlation for a building's envelope in a Mediterranean climate in Morocco'. Int J Renewable Energy Res 11: 1064–1074.
[8]	Ozarisoy B, Altan H (2022) Significance of occupancy patterns and habitual household adaptive behaviour on home-energy performance of post-war social-housing estate in the south-eastern Mediterranean climate: Energy policy design. Energy 244: 122904. https://doi.org/10.1016/j.energy.2021.122904 doi: 10.1016/j.energy.2021.122904
[9]	Dai B, Wang Q, Liu S, et al. (2023) Novel configuration of dual-temperature condensation and dual-temperature evaporation high-temperature heat pump system: Carbon footprint, energy consumption, and financial assessment. Energy Convers Manage 292: 117360. https://doi.org/10.1016/j.enconman.2023.117360 doi: 10.1016/j.enconman.2023.117360
[10]	Miyata S, Lim J, Akashi Y, et al. (2020) Optimal set-point regulation in HVAC system for controllability and energy efficiency. Adv Build Energy Res 14: 160–170. https://doi.org/10.1080/17512549.2019.1578263 doi: 10.1080/17512549.2019.1578263
[11]	Gao G, Li J, Wen Y (2020) DeepComfort: Energy-efficient thermal comfort control in buildings via reinforcement learning. IEEE Internet Things 7: 8472–8484. https://doi.org/10.1109/JIOT.2020.2992117 doi: 10.1109/JIOT.2020.2992117
[12]	Zhang Z, Chong A, Pan Y, et al. (2019) Whole building energy model for HVAC optimal control: A practical framework based on deep reinforcement learning. Energy Build 199: 472–490. https://doi.org/10.1016/j.enbuild.2019.07.029 doi: 10.1016/j.enbuild.2019.07.029
[13]	Jin Y, Yan D, Zhang X, et al. (2021) A data-driven model predictive control for lighting system based on historical occupancy in an office building: Methodology development. Build Simul 14: 219–235. https://doi.org/10.1007/s12273-020-0638-x doi: 10.1007/s12273-020-0638-x
[14]	Awang M, Tham CS, Ruddin NMB, et al. (2020) Assessment of energy saving potential and lighting system in teaching building. J Adv Res Fluid Mech Therm Sci 65: 159–169.
[15]	Yan K, Li W, Ji Z, et al. (2019) A hybrid LSTM neural network for energy consumption forecasting of individual households. IEEE Access 7: 157633–157642. https://doi.org/10.1109/ACCESS.2019.2949065 doi: 10.1109/ACCESS.2019.2949065
[16]	Silva J, Praça I, Pinto T, et al. (2020) Energy consumption forecasting using ensemble learning algorithms. In: Herrera-Viedma, E., Vale, Z., Nielsen, P., Martin Del Rey, A., Casado Vara, R., Distributed Computing and Artificial Intelligence, 16th International Conference, Special Sessions. DCAI 2019. Advances in Intelligent Systems and Computing, Springer: Cham, 5–13. https://doi.org/10.1007/978-3-030-23946-6_1
[17]	Wei N, Li C, Peng X, et al. (2019) Conventional models and artificial intelligence-based models for energy consumption forecasting: A review. J Pet Sci Eng 181: 106187. https://doi.org/10.1016/j.petrol.2019.106187 doi: 10.1016/j.petrol.2019.106187
[18]	Wang R, Lu S, Feng W (2020) A novel improved model for building energy consumption prediction based on model integration. Appl Energy 262: 114561. https://doi.org/10.1016/j.apenergy.2020.114561 doi: 10.1016/j.apenergy.2020.114561
[19]	Daut MAM, Hassan MY, Abdullah H, et al. (2017) Building electrical energy consumption forecasting analysis using conventional and artificial intelligence methods: A review. Renewable Sustainable Energy Rev 70: 1108–1118. https://doi.org/10.1016/j.rser.2016.12.015 doi: 10.1016/j.rser.2016.12.015
[20]	Chou JS, Bui DK (2014) Modeling heating and cooling loads by artificial intelligence for energy-efficient building design. Energy Build 82: 437–446. https://doi.org/10.1016/j.enbuild.2014.07.036 doi: 10.1016/j.enbuild.2014.07.036
[21]	Newsham GR, Birt BJ (2010) Building-level occupancy data to improve ARIMA-based electricity use forecasts. Proceedings of the 2nd ACM Workshop on Embedded Sensing Systems for Energy-Efficiency in Building, 13–18. https://doi.org/10.1145/1878431.1878435 doi: 10.1145/1878431.1878435
[22]	Bilgili M, Pinar E (2023) Gross electricity consumption forecasting using LSTM and SARIMA approaches: A case study of Türkiye. Energy 284: 128575. https://doi.org/10.1016/j.energy.2023.128575 doi: 10.1016/j.energy.2023.128575
[23]	Blázquez-García A, Conde A, Milo A, et al. (2020) Short-term office building elevator energy consumption forecast using SARIMA. J Build Perform Simul 13: 69–78. https://doi.org/10.1080/19401493.2019.1698657 doi: 10.1080/19401493.2019.1698657
[24]	Wang X, Meng M (2012) A hybrid neural network and ARIMA model for energy consumption forcasting. J Comput 7: 1184–1190. https://doi.org/10.4304/jcp.7.5.1184-1190 doi: 10.4304/jcp.7.5.1184-1190
[25]	Kandananond K (2011) Forecasting electricity demand in Thailand with an artificial neural network approach. Energies 4: 1246–1257. https://doi.org/10.3390/en4081246 doi: 10.3390/en4081246
[26]	Wang JQ, Du Y, Wang J (2020) LSTM based long-term energy consumption prediction with periodicity. Energy 197: 117197. https://doi.org/10.1016/j.energy.2020.117197 doi: 10.1016/j.energy.2020.117197
[27]	Chujai P, Kerdprasop N, Kerdprasop K (2013) Time series analysis of household electric consumption with ARIMA and ARMA models. Proceedings of the International Multiconference of Engineers and Computer Scientists, Hong Kong: IAENG, 1: 295–300.
[28]	Liu J, Zhao Z, Zhong Y, et al. (2022) Prediction of the dissolved gas concentration in power transformer oil based on SARIMA model. Energy Rep 8: 1360–1367. https://doi.org/10.1016/j.egyr.2022.03.020 doi: 10.1016/j.egyr.2022.03.020
[29]	Camara A, Feixing W, Xiuqin L (2016) Energy consumption forecasting using seasonal ARIMA with artificial neural networks models. Int J Bus Manage 11: 231–243. https://doi.org/10.5539/ijbm.v11n5p231 doi: 10.5539/ijbm.v11n5p231
[30]	Dhakal S, Gautam Y, Bhattarai A (2021) Exploring a deep LSTM neural network to forecast daily PM 2.5 concentration using meteorological parameters in Kathmandu Valley, Nepal. Air Qual Atmos Health 14: 83–96. https://doi.org/10.1007/s11869-020-00915-6 doi: 10.1007/s11869-020-00915-6
[31]	Khan ZA, Hussain T, Ullah A, et al. (2020) Towards efficient electricity forecasting in residential and commercial buildings: A novel hybrid CNN with a LSTM-AE based framework. Sensors 20: 1399. https://doi.org/10.3390/s20051399 doi: 10.3390/s20051399
[32]	Kim EG, Akhtar MS, Yang OB (2023) Designing solar power generation output forecasting methods using time series algorithms. Electr Power Syst Res 216: 109073. https://doi.org/10.1016/j.epsr.2022.109073 doi: 10.1016/j.epsr.2022.109073
[33]	Siami-Namini S, Tavakoli N, Namin AS (2018) A comparison of ARIMA and LSTM in forecasting time series. In 2018 17th IEEE International Conference on Machine Learning and Applications (ICMLA), IEEE, 1394–1401. https://doi.org/10.1109/ICMLA.2018.00227

This article has been cited by:

1.	S.A. Pedro, J.M. Tchuenche, HIV/AIDS dynamics: Impact of economic classes with transmission from poor clinical settings, 2010, 267, 00225193, 471, 10.1016/j.jtbi.2010.09.019
2.	C. P. Bhunu, S. Mushayabasa, H. Kojouharov, J. M. Tchuenche, Mathematical Analysis of an HIV/AIDS Model: Impact of Educational Programs and Abstinence in Sub-Saharan Africa, 2011, 10, 1570-1166, 31, 10.1007/s10852-010-9134-0
3.	Sarada S. Panchanathan, Diana B. Petitti, Douglas B. Fridsma, The development and validation of a simulation tool for health policy decision making, 2010, 43, 15320464, 602, 10.1016/j.jbi.2010.03.013
4.	Cristiana J. Silva, Delfim F.M. Torres, A SICA compartmental model in epidemiology with application to HIV/AIDS in Cape Verde, 2017, 30, 1476945X, 70, 10.1016/j.ecocom.2016.12.001
5.	S. D. Hove-Musekwa, F. Nyabadza, H. Mambili-Mamboundou, Modelling Hospitalization, Home-Based Care, and Individual Withdrawal for People Living with HIV/AIDS in High Prevalence Settings, 2011, 73, 0092-8240, 2888, 10.1007/s11538-011-9651-7
6.	Emmanuelina L. Kateme, Jean M. Tchuenche, Senelani D. Hove-Musekwa, HIV/AIDS Dynamics with Three Control Strategies: The Role of Incidence Function, 2012, 2012, 2090-5572, 1, 10.5402/2012/864795
7.	F. Nyabadza, Z. Mukandavire, S.D. Hove-Musekwa, Modelling the HIV/AIDS epidemic trends in South Africa: Insights from a simple mathematical model, 2011, 12, 14681218, 2091, 10.1016/j.nonrwa.2010.12.024
8.	Xiaodong Wang, Chunxia Wang, Kai Wang, Extinction and persistence of a stochastic SICA epidemic model with standard incidence rate for HIV transmission, 2021, 2021, 1687-1847, 10.1186/s13662-021-03392-y

Reader Comments

Your name:*

Email:*
© 2023 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)