
Citation: Yanxiang Tan. Dynamics analysis of Mackey-Glass model with two variable delays[J]. Mathematical Biosciences and Engineering, 2020, 17(5): 4513-4526. doi: 10.3934/mbe.2020249
[1] | A. Vinodkumar, T. Senthilkumar, S. Hariharan, J. Alzabut . Exponential stabilization of fixed and random time impulsive delay differential system with applications. Mathematical Biosciences and Engineering, 2021, 18(3): 2384-2400. doi: 10.3934/mbe.2021121 |
[2] | Gonzalo Robledo . Feedback stabilization for a chemostat with delayed output. Mathematical Biosciences and Engineering, 2009, 6(3): 629-647. doi: 10.3934/mbe.2009.6.629 |
[3] | Fawaz E Alsaadi, Chuangxia Huang, Madini O Alassafi, Reem M Alotaibi, Adil M Ahmad, Jinde Cao . Attractivity criterion on a delayed tick population dynamics equation with a reproductive function $ f(u) = ru^{\gamma}e^{-\sigma u} $. Mathematical Biosciences and Engineering, 2022, 19(12): 12852-12865. doi: 10.3934/mbe.2022600 |
[4] | Yuanpei Xia, Weisong Zhou, Zhichun Yang . Global analysis and optimal harvesting for a hybrid stochastic phytoplankton-zooplankton-fish model with distributed delays. Mathematical Biosciences and Engineering, 2020, 17(5): 6149-6180. doi: 10.3934/mbe.2020326 |
[5] | Zhongcai Zhu, Xiaomei Feng, Xue He, Hongpeng Guo . Mirrored dynamics of a wild mosquito population suppression model with Ricker-type survival probability and time delay. Mathematical Biosciences and Engineering, 2024, 21(2): 1884-1898. doi: 10.3934/mbe.2024083 |
[6] | Salih Djillali, Soufiane Bentout, Tarik Mohammed Touaoula, Abdessamad Tridane . Global dynamics of alcoholism epidemic model with distributed delays. Mathematical Biosciences and Engineering, 2021, 18(6): 8245-8256. doi: 10.3934/mbe.2021409 |
[7] | Mostafa Adimy, Abdennasser Chekroun, Claudia Pio Ferreira . Global dynamics of a differential-difference system: a case of Kermack-McKendrick SIR model with age-structured protection phase. Mathematical Biosciences and Engineering, 2020, 17(2): 1329-1354. doi: 10.3934/mbe.2020067 |
[8] | Farzad Fatehi, Yuliya N. Kyrychko, Konstantin B. Blyuss . Time-delayed model of autoimmune dynamics. Mathematical Biosciences and Engineering, 2019, 16(5): 5613-5639. doi: 10.3934/mbe.2019279 |
[9] | Jiawei Deng, Ping Jiang, Hongying Shu . Viral infection dynamics with mitosis, intracellular delays and immune response. Mathematical Biosciences and Engineering, 2023, 20(2): 2937-2963. doi: 10.3934/mbe.2023139 |
[10] | Masaki Sekiguchi, Emiko Ishiwata, Yukihiko Nakata . Dynamics of an ultra-discrete SIR epidemic model with time delay. Mathematical Biosciences and Engineering, 2018, 15(3): 653-666. doi: 10.3934/mbe.2018029 |
The complex characterization of delay dynamic model has been the hot topic in diverse applications of physics, mathematical biology, economic finance and networks systems, etc. [1,2,3,4,5,6,7,8,9,10,11,12,13,14]. As early as 1977, for the sake of investigating the regulation and control mechanisms of human diseases, Mackey and Glass [15] put forward the following famous delay differential equation:
dP(t)dt=−γP(t)+β0θnP(t−τ)θn+Pn(t−τ) | (1.1) |
to describe hematopoiesis with a single-humped production rate, which is now called as Mackey-Glass model. Biologically, P(t) is the density of mature circulating cells, n>0, γ,β0,θ are positive constants and denote the destruction rate, maximal production rate, shape parameter respectively, τ represents the time delay from immature to mature cells produced by bone marrow in the circulations. The simple mathematical model of physiological system (1.1) can reproduce certain qualitative features of normal and abnormal respiration, and the existence of positive solutions, bifurcation, permanence, boundedness, oscillation, and local stability of such equation and similar models have greatly attracted the attention of both theoretical and empirical communities (for example [16,17,18,19,20,21]).
In order to analyze the effects of maturity time and incubation time on system dynamics, system (1.1) was generalized to the following nonautonomous delay differential equation form [20]
dx(t)dt=β(t)[ax(t−h(t))1+xν(t−g(t))−x(t)], a>0, ν>0,t≥t0, | (1.2) |
in which g(t) represents the developmental or maturation time, h(t) represents the incubation time. It should be assumed that β,g, h:[t0, +∞)→(0, +∞) are continuous functions with
0<β−=inft∈[t0, +∞)β(t)≤supt∈[t0, +∞)β(t)=β+<+∞, | (1.3) |
and
τ:=max{τ1,τ2}≥r:=min{r1, r2}>0, | (1.4) |
where
τ1=supt∈[t0, +∞)h(t),τ2=supt∈[t0, +∞)g(t),r1=inft∈[t0, +∞)h(t),r2=inft∈[t0, +∞)g(t). |
Just as shown in [20], the introduction of two different time delays (g(t) and h(t)) in the model shows more complex dynamic behavior than the previous single delay (τ). For instance, practical examples with two delays instead of single delay can produce sustained oscillations or even chaotic phenomenon [22,23,24,25,26,27,28,29]. Dynamics on the famous Mackey-Glass model have not been further explored yet in two variable delays case, which was first proposed by Berezansky and Braverman as open problems [20]. In fact, when two delays occur in system (1.2), the delayed feedback function can only be treated as a binary function, which brings many essential research difficulties. Up to now, the investigation on the global stability of the model (1.2) is still a challenging open problem, and the research results are very rare, we only find [20] in the existing work, however, the authors of [20] assume that a>1 and put some additional technical conditions on the delay terms. More precisely, one of the main results of [20] is established as follows.
Theorem 1.1 (see Theorem 2.1 in [20]) Suppose that a>1, ∫∞0β(t)dt=+∞, and x(t) is a positive global solution of system (1.2). Then system (1.2) has the unique positive equilibrium N∗=(a−1)1ν. Moreover,
(ⅰ) x(t)>N∗,t≥t0, implies N∗=lim inft→∞x(t);
(ⅱ) x(t)∈(0,N∗),t≥t0 yields that N∗=lim supt→∞x(t).
Now, a natural question is: whether the non-oscillatory solutions mentioned in Theorem 1.1 converge to N∗. Sparked by the above reason and discussion, we try to give a satisfying answer and establish the global asymptotical stability of the Mackey-Glass model (1.2) with two different delays h≢g, a>1 or 0<a≤1.
The rest of the proposed work is furnished as follows: In section 2, some necessary symbols and preliminary results are introduced. Two delay-independent criteria based on the fluctuation lemma and techniques of differential inequality are established in section 3. One numerical example and its computer simulations are provided to illustrate the effectiveness of the acquired results in section 4. At last, section 5 contains conclusions.
Before stating our results, we present some symbols. Let C=C([−τ, 0],R) be the Banach space of the set of all continuous functions from [−τ, 0] to R equipped with the supremun norm ||⋅|| and C+={φ∈C| φ(θ)≥0 for θ∈[−τ, 0]}. Define xt∈C by xt(θ)=x(t+θ) for θ∈[−τ, 0], where x is a continuous function defined on [−τ, σ) with 0≤t<σ. The initial condition with respect to system (1.2) is as below
xt0=φ, φ∈C+. | (2.1) |
Lemma 2.1. For φ∈C+, the solution of system (1.2) with respect to φ can be denoted by x(t;t0,φ), then x(t;t0,φ) is nonnegative, unique and exists on [t0, +∞). Moreover, for φ∈C+ and φ(0)>0, x(t;t0,φ)>0 for all t∈[t0, +∞).
Proof. By using the method of step by step, one can show that x(t)=x(t;t0,φ) is unique and exists on [t0, +∞). As φ∈C+, according to systems (1.2) and (1.4), we obtain
x(t)=φ(0)e−∫tt0β(v)dv+e−∫tt0β(v)dv∫tt0β(s)aφ(s−h(s)−t0)1+φν(s−g(s)−t0)e∫st0β(v)dvds≥0, t∈[t0, t0+r]. |
By the method of steps, we have x(t)≥0 on [t0+r, t0+2r],[t0+2r, t0+3r]⋯,[t0+nr, t0+(n+1)r],⋯. Thus, x(t)≥0 for all t≥t0.
Moreover, if φ∈C+ and φ(0)>0, then
x(t)≥φ(0)e−∫tt0β(v)dv+ae−∫tt0β(v)dv∫tt0β(s)x(s−h(s))1+xν(s−g(s))e∫st0β(v)dvds>0, |
for all t∈[t0, +∞). This ends the proof.
Lemma 2.2. Let 0<a≤1 and x(t)=x(t;t0,φ), if φ∈C+, then for all t∈[t0, +∞), we have x(t)≤‖φ‖.
Proof. For φ∈C+, we state that x(t)=x(t;φ)≤‖φ‖ for all t∈[t0, t0+r]. Otherwise, there is t1∈(t0, t0+r] obeying
x′(t1)>0, x(t1)>‖φ‖, x(s)≤‖φ‖,s∈[t0,t1). | (2.2) |
Furthermore, systems (1.2) and (2.2) lead to
0<x′(t1)=β(t1)[ax(t1−h(t1))1+xν(t1−g(t1))−x(t1)]<β(t1)[ax(t1−h(t1))−‖φ‖]≤β(t1)(a−1)‖φ‖≤0. |
This is a clear contradiction and reveals the above statement. Then, x(t)≤‖φ‖ holds on [t0, +∞) by induction. The proof of Lemma 2.2 is now completed.
Remark 2.1. Suppose that 0<a≤1 is satisfied. For any ϵ>0, we can choose ˜δ∈[0, ϵ). Then, for any φ∈C+ with ‖φ‖<˜δ, from Lemma 2.2, we can show that
x(t)=x(t;t0,φ)≤‖φ‖<˜δ<ϵ for all t∈[t0,+∞). |
This suggests that 0 is locally stable for Eq (1.2) with 0<a≤1.
First, we establish two delay-independent criteria to show the attractivity of all non-oscillating solutions about the positive equilibrium point.
Theorem 3.1. Let a>1, x(t)=x(t;t0,φ) with φ∈C+ and φ(0)>0, and there exists T1>t0 such that x(t)≤N∗ for all t≥T1. Then limt→+∞x(t)=N∗.
Proof. From Lemma 2.1, it is obvious that 0<x(t)≤N∗ for all t≥T1+τ. We claim that l:=lim inft→+∞x(t)>0. Otherwise, by the definition of l, we have l=0. For each t≥T1+τ, we define
m(t)=min{ξ∈[T1+τ, t]:x(ξ)=minT1+τ≤s≤tx(s)}. |
It follows from l=0 that limt→+∞m(t)=+∞, limt→+∞x(m(t))=0, and
0≥x′(m(t)), x(m(t)−h(m(t)))>x(m(t))>0, for all m(t)>T1+2τ, |
which, together with (1.2), implies that
0≥x′(m(t))=β(m(t))[ax(m(t)−h(m(t)))1+xν(m(t)−g(m(t)))−x(m(t))]>β(m(t))x(m(t))[a1+(N∗)ν−1]=0, where m(t)>T1+2τ. |
This is a contradiction and suggests that l>0.
By the fluctuation lemma (see Lemma A.1 in [30]), one can pick a sequence {tk}k≥1 agreeing with
tk→+∞,x(tk)→lim inft→+∞x(t)=l,x′(tk)→0as k→+∞. |
From the boundedness on β(⋅), we can select a subsequence of {β(tk)}k≥1, still denoted by {β(tk)}k≥1, such that limk→+∞β(tk) exists and limk→+∞β(tk)≥β−>0. This follows that
0=limk→+∞x′(tk)=limk→+∞β(tk)limk→+∞[ax(tk−h(tk))1+xν(tk−g(tk))−x(tk)]≥l[a1+(lim supk→+∞x(tk−g(tk)))ν−1]limk→+∞β(tk) |
and then lim supk→+∞x(tk−g(tk))=N∗, which, implies that lim supt→+∞x(t)=N∗ since lim supt→+∞x(t)≤N∗.
For any ε∈(0, lim inft→+∞x(t)), we select ˜T>T1+2τ, such that for all t≥˜T, we have
x(t)>lim inft→+∞x(t)−ε. |
Consequently, for all t≥˜T+τ, one have
(x(t)e∫tt0β(v)dv)′=e∫tt0β(v)dv[x′(t)+β(t)x(t)]=e∫tt0β(v)dvaβ(t)x(t−h(t))1+xν(t−g(t))>e∫tt0β(v)dvaβ(t)[lim inft→+∞x(t)−ε]11+(N∗)ν=e∫tt0β(v)dvβ(t)[lim inft→+∞x(t)−ε]. | (3.1) |
Take k0>0 satisfying k0τ1≥˜T+τ for all k≥k0. Given a positive integer k≥k0, we denote
x(Sk)=maxt∈[kτ1,(k+1)τ1]x(t), x(sk)=mint∈[kτ1,(k+1)τ1]x(t), where sk, Sk∈[kτ1,(k+1)τ1]. |
Now, we state that {x(sk)}k≥k0 is nondecreasing. Suppose the contrary and choose a positive integer n0≥k0 with x(sn0)>x(sn0+1). Then, sn0+1∈((n0+1)τ1,(n0+2)τ1], and there is s∗=inf{ω∈((n0+1)τ1,(n0+2)τ1]:x(ω)=x(sn0+1)} agreeing with
x′(s∗)≤0, and x(s∗)<x(t) for all t∈[n0τ1, s∗). |
Subsequently, system (1.2) gives us that
0≥x′(s∗)=β(s∗)[ax(s∗−h(s∗))1+xν(s∗−g(s∗))−x(s∗)]>β(s∗)[x(s∗)(a1+(N∗)ν−1)]=0, | (3.2) |
is a contradiction. This validates the above statement, which implies that
limk→+∞x(sk)=lim inft→+∞x(t). | (3.3) |
Take a subsequence {kj}j≥1 of {k}k≥k0 such that
limj→+∞x(Skj)=lim supt→+∞x(t)=N∗. | (3.4) |
With the aid of the boundedness of {∫Sksk+1β(v)dv}k≥k0, we can select its subsequence, still labeled by {∫Sksk+1β(v)dv}k≥k0, such that limj→+∞∫Skjskj+1β(v)dv exists and
limj→+∞∫Skjskj+1β(v)dv=τ∗∈[−2τ1β+, 0]. | (3.5) |
With the help of an integration of system (3.1) on [Skj, skj+1], we see that
x(skj+1)e∫skj+1t0β(v)dv≥x(Skj)e∫Skjt0β(v)dv+[lim inft→+∞x(t)−ε]∫skj+1Skje∫tt0β(v)dvβ(t)dt=x(Skj)e∫Skjt0β(v)dv+[lim inft→+∞x(t)−ε](e∫skj+1t0β(v)dv−e∫Skjt0β(v)dv) |
and
x(skj+1)≥x(Skj)e∫Skjskj+1β(v)dv+[lim inft→+∞x(t)−ε][1−e∫Skjskj+1β(v)dv]. | (3.6) |
According to Eqs (3.3)–(3.5) and inequality (3.6), setting j→+∞ produces
lim inft→+∞x(t)≥N∗eτ∗+[lim inft→+∞x(t)−ε][1−eτ∗]. |
It follows from the arbitrariness of ε that lim inft→+∞x(t)=N∗, and proves Theorem 3.1.
Theorem 3.2. Let a>1, x(t)=x(t;t0,φ) with φ∈C+, and there exists T2>t0 such that x(t)≥N∗ for all t≥T2. Then limt→+∞x(t)=N∗.
Proof. We state that lim supt→+∞x(t)<+∞. Otherwise, lim supt→+∞x(t)=+∞. For each t≥T2+τ, we label
˜M(t)=min{ξ∈[T2+τ, t]:x(ξ)=maxT2+τ≤s≤tx(s)}. |
It follows from lim supt→+∞x(t)=+∞ that limt→+∞˜M(t)=+∞, limt→+∞x(˜M(t))=+∞, and
x(˜M(t))>x(˜M(t)−h(˜M(t))), for all ˜M(t)>T2+2τ. |
According to the definition of ˜M(t), we know that
0≤x′(˜M(t))=β(˜M(t))[ax(˜M(t)−h(˜M(t)))1+xν(˜M(t)−g(˜M(t)))−x(˜M(t))]<β(˜M(t))x(˜M(t))(a1+(N∗)ν−1)=0, where ˜M(t)>T2+2τ. |
This is a contradiction and suggests that lim supt→+∞x(t)<+∞. Using the fluctuation lemma (see Lemma A.1 in [30]), and considering the boundedness of β(t), one can find a sequence {t∗k}k≥1 obeying that
t∗k→+∞,x(t∗k)→lim supt→+∞x(t)=L,x′(t∗k)→0as k→+∞, |
and limk→+∞β(t∗k)≥β−>0. This entails that
0=limk→+∞x′(t∗k)=limk→+∞β(t∗k)limk→+∞[ax(t∗k−h(t∗k)))1+xν(t∗k−g(t∗k)−x(t∗k)]≤limk→+∞β(t∗k)L[a1+(lim infk→+∞x(t∗k−g(t∗k)))ν−1] |
and then lim infk→+∞x(t∗k−g(t∗k))=N∗. This and the fact that lim inft→+∞x(t)≥N∗ produce lim inft→+∞x(t)=N∗.
For any ε>0, one can pick ˜T1>T2 agreeing with
x(t)<lim supt→+∞x(t)+ε for all t≥˜T1. |
Furthermore, for all t≥˜T1+τ, we gain
(x(t)e∫tt0β(v)dv)′=e∫tt0β(v)dv[x′(t)+β(t)x(t)]=e∫tt0β(v)dvβ(t)ax(t−h(t))1+xν(t−g(t))<e∫tt0β(v)dvβ(t)a1+(N∗)ν[lim supt→+∞x(t)+ε]=e∫tt0β(v)dvβ(t)[lim supt→+∞x(t)+ε]. | (3.7) |
Set k∗0>0 with k∗0τ1≥˜T1+τ for all k≥k∗0. Given a positive integer k obeying k≥k∗0, we denote
x(S∗k)=maxt∈[kτ1,(k+1)τ1]x(t), x(s∗k)=mint∈[kτ1,(k+1)τ1]x(t), |
where s∗k, S∗k∈[kτ1,(k+1)τ1]. Hereafter, we assert that {x(S∗k)}k≥k∗0 is non-increasing. Suppose that, on the contrary, there is a positive integer n∗0≥k∗0 obeying x(S∗n∗0)<x(S∗n∗0+1). Then, Sn∗0+1∈((n∗0+1)τ1,(n∗0+2)τ1], and we can find S∗=inf{ω∈((n∗0+1)τ1,(n∗0+2)τ1]:x(ω)=x(S∗n∗0+1)} agreeing with
x′(S∗)≥0, and x(S∗)>x(t) for all t∈[n∗0τ1,S∗). |
It follows from system (1.2) that
0≤x′(S∗)=β(S∗)[ax(S∗−h(S∗))1+xν(S∗−g(S∗))−x(S∗)]<β(S∗)x(S∗)(a1+(N∗)ν−1)=0, | (3.8) |
which is a contradiction. This furnishes the above assertion and results that
limk→+∞x(S∗k)=lim supt→+∞x(t). | (3.9) |
Now, from the definition of {s∗k}k≥k∗0 and the boundedness of {∫s∗kS∗k+1β(v)dv}k≥k∗0, we can pick a subsequence {kj}j≥1 of {k}k≥k∗0 such that
limj→+∞x(s∗kj)=lim inft→+∞x(t)=N∗, | (3.10) |
and
limj→+∞∫s∗kjS∗kj+1β(v)dv=τ∗∗∈[−2τ1β+, 0]. | (3.11) |
Computing the integration of (3.7) on [s∗kj, S∗kj+1] yields
x(S∗kj+1)e∫S∗kj+1t0β(v)dv≤x(s∗kj)e∫s∗kjt0β(v)dv+[lim supt→+∞x(t)+ε]∫S∗kj+1s∗kje∫tt0β(v)dvβ(t)dt=x(s∗kj)e∫s∗kjt0β(v)dv+[lim supt→+∞x(t)+ε](e∫S∗kj+1t0β(v)dv−e∫s∗kjt0β(v)dv) |
and
x(S∗kj+1)<x(s∗kj)e∫s∗kjS∗kj+1β(v)dv+[lim supt→+∞x(t)+ε][1−e∫s∗kjS∗kj+1β(v)dv]. | (3.12) |
In view of (3.9), (3.10), (3.11) and (3.12), we gain
lim supt→+∞x(t)≤N∗eτ∗∗+[lim supt→+∞x(t)+ε][1−eτ∗∗], |
which produces that lim supt→+∞x(t)=N∗ and reveals Theorem 3.2.
Next, we establish the global asymptotical stability on the zero equilibrium point. More precisely,
Theorem 3.3. If 0<a≤1, then for system (1.2), the trivial solution 0 is globally asymptotically stable with respect to C+.
Proof. From Lemma 2.2 and Remark 2.1, it is obvious that x(t)=x(t;t0,φ) with φ∈C+ is bounded, and 0 is a local stable. In order to prove Theorem 3.3, it is sufficient to show that lim supt→+∞x(t)=0. Suppose the contrary and let
lim supt→+∞x(t)=L>0. | (3.13) |
Using the fluctuation lemma (see Lemma A.1 in [30]), and considering the boundedness of β(t), one can find a sequence {T∗k}k≥1 with
T∗k→+∞,x(T∗k)→lim supt→+∞x(t)=L,x′(T∗k)→0as k→+∞, |
and limk→+∞β(T∗k)≥β−>0. This leads to
0=limk→+∞x′(T∗k)=limk→+∞β(T∗k)limk→+∞[ax(T∗k−h(T∗k)))1+xν(T∗k−g(T∗k)−x(T∗k)]≤limk→+∞β(T∗k)L[a1+(lim infk→+∞x(T∗k−g(T∗k)))ν−1] |
and then
lim infk→+∞x(T∗k−g(T∗k))=0=lim inft→+∞x(t). |
Given ε>0, we can take ˆT>t0+τ agreeing with
x(t)<L+ε for all t≥ˆT. |
Then, 0<a≤1 produces
(x(t)e∫tt0β(v)dv)′=e∫tt0β(v)dv[x′(t)+β(t)x(t)]=e∫tt0β(v)dvβ(t)ax(t−h(t))1+xν(t−g(t))≤e∫tt0β(v)dvβ(t)(L+ε), for all t≥ˆT+τ. | (3.14) |
Choose k∗∗0>0 satisfying k∗0τ1≥ˆT+τ for all k≥k∗∗0, for k≥k∗∗0, we denote
x(Qk)=maxt∈[kτ1,(k+1)τ1]x(t), x(qk)=mint∈[kτ1,(k+1)τ1]x(t), |
where qk, Qk∈[kτ1,(k+1)τ1]. Furthermore, we state that {x(Qk)}k≥k∗0 is non-increasing. If the statement is not true, then we can choose a positive integer n∗∗0≥k∗∗0, such that x(Qn∗∗0)<x(Qn∗∗0+1). Consequently, we get Qn∗∗0+1∈((n∗∗0+1)τ1,(n∗∗0+2)τ1], and there exists Q∗=inf{ω∈((n∗∗0+1)τ1,(n∗∗0+2)τ1]:x(Qn∗∗0+1)=x(ω)} satisfying
x′(Q∗)≥0, and x(Q∗)>x(t) for all t∈[n0τ1, Q∗). |
However, system (1.2) entails that
0≤x′(Q∗)=β(Q∗)[ax(Q∗−h(Q∗))1+xν(Q∗−g(Q∗))−x(Q∗)]<β(Q∗)x(Q∗)(a1−1)≤0, | (3.15) |
which is a contradiction. This results the above statement and implies that
limk→+∞x(Qk)=L. | (3.16) |
Now, making use of the definition of {qk}k≥k∗∗0 and by employing the boundedness of {e∫qkQk+1β(v)dv}k≥k∗∗0, it is easy to see that there is a subsequence {kj}j≥1 of {k}k≥k∗∗0 such that
limj→+∞x(qkj)=lim inft→+∞x(t)=0,limj→+∞e∫qkjQkj+1β(v)dv=Q∗∗∈[−2τ1β+, 0]. | (3.17) |
From inequality (3.14), we obtain
x(Qkj+1)e∫Qkj+1t0β(v)dv≤x(qkj)e∫qkjt0β(v)dv+(L+ε)∫Qkj+1qkje∫tt0β(v)dvβ(t)dt=x(qkj)e∫qkjt0β(v)dv+(L+ε)(e∫Qkj+1t0β(v)dv−e∫qkjt0β(v)dv), |
and
x(Qkj+1)<x(qkj)e∫qkjQkj+1β(v)dv+(L+ε)[1−e∫qkjQkj+1β(v)dv]. | (3.18) |
With the aid of Eqs (3.16) and (3.17) and inequality (3.18), setting j→+∞ yields
L≤(L+ε)(1−eQ∗∗), |
which entails that lim supt→+∞x(t)=L=0. This contradicts Eq (3.13) and proves Theorem 3.3.
Remark 3.1. Clearly, under assumption (1.4), all results in Theorem 1.1 are special ones of Theorems 3.1 and 3.2 in this present paper. In particular, we can show that the two different time-varying delays h(t) and g(t) do not yield the effect on the asymptotic behavior on the non-oscillating solutions about the positive equilibrium point of Mackey-Glass equation. Also, Theorem 3.3 implies that, for 0<a≤1, the global stability of non-autonomous Mackey-Glass equation is independent of the two different time-varying delays h(t) and g(t).
In this section, we present two examples with graphical illustration to show the theoretical findings in this paper. Consider the following delay Mackey-Glass equations:
x′(t)=(10+sint)[28x(t−(10+cost))1+x3(t−(30+cos2t))−x(t)], t≥t0=0, | (4.1) |
and
x′(t)=e2sin3t[x(t−(10+cost))1+x3(t−(30+cos2t))−x(t)], t≥t0=0. | (4.2) |
From Theorems 3.1 and 3.2, one can see that all non-oscillating solutions about the positive equilibrium point N∗=3 of Eq (4.1) are convergent to N∗. Moreover, Theorem 3.3 suggests that Eq (4.2) is globally asymptotically stable. Computer simulations of the two numerical examples are arranged in Figures 1 and 2 to illustrate the correctness and effectiveness of the presented convergence results.
Remark 4.1. We should point out that the delay-independent criteria for the asymptotic behavior of the non-autonomous Mackey-Glass equation with two different delays have not been established in the previous literature. Most recently, the authors in [31] used the classical approach of "decomposing + embedding" to investigate the global attraction for the autonomous equation
x′(t)=βx(t)+βF(x(t−σ),x(t−τ)), |
in the case that β,σ,τ∈[0, +∞) are constants and σ≤τ. On the other hand, I. Győri, et al. [32] studied the permanence of the below system
x′(t)=α(t)H(x(t−σ),x(t−τ))−β(t)f(x(t)). |
Obviously, all results in [17,18,19,20,31,32,33,34] cannot be applicable to show convergence of the solutions of Eqs (4.1) and (4.2).
In this paper, we mainly investigate the attractivity and the global asymptotical stability for a class of non-autonomous Mackey-Glass model with two variable delays. Firstly, in the case of a unique positive equilibrium point, we prove that two different variable delays appearing in Mackey-Glass equation do not yield the effect on the asymptotic behavior of the non-oscillating solutions about the positive equilibrium point. Then two delay-independent criteria based on the fluctuation lemma and techniques of differential inequality are established. The obtained results improve and complement some published results, and partly answer the open problem proposed by Berezansky and Braverman [20]. By the way, the approach proposed in this article furnishes a possible method for studying the asymptotic behavior on the other population dynamic models involving the different reproductive function delays.
The author would like to express the sincere appreciation to the editor and reviewers for their helpful comments in improving the presentation and quality of the paper. This work was supported by the National Natural Science Foundation of China (No. 11971076), Research Promotion Program of Changsha University of Science and Technology (No. 2019QJCZ050).
The author confirms that he has no conflict of interest.
[1] | H. Hu, X. Yuan, L. Huang, C. Huang, Global dynamics of an SIRS model with demographics and transfer from infectious to susceptible on heterogeneous networks, Math. Biosci. Eng., 16 (2019), 5729-5749. |
[2] | C. Qian, Y. Hu, Novel stability criteria on nonlinear density-dependent mortality Nicholson's blowflies systems in asymptotically almost periodic environments, J. Inequal. Appl., 2020 (2020), 13. |
[3] | C. Huang, H. Zhang, L. Huang, Almost periodicity analysis for a delayed Nicholson's blowflies model with nonlinear density-dependent mortality term, Commun. Pure Appl. Anal., 18 (2019), 3337-3349. |
[4] | Y. Tan, C. Huang, B. Sun, T. Wang, Dynamics of a class of delayed reaction-diffusion systems with Neumann boundary condition, J. Math. Anal. Appl., 458 (2018), 1115-1130. |
[5] | C. Huang, X. Long, L. Huang, S. Fu, Stability of almost periodic Nicholson's blowflies model involving patch structure and mortality terms, Canad. Math. Bull., 63 (2020), 405-422. |
[6] | C. Huang, H. Zhang, J. Cao, H. Hu, Stability and Hopf bifurcation of a delayed prey-predator model with disease in the predator, Internat. J. Bifur. Chaos Appl., 29 (2019), 1950091. |
[7] | X. Long, S. Gong, New results on stability of Nicholson's blowflies equation with multiple pairs of time-varying delays, Appl. Math. Lett., 100 (2020), 106027. |
[8] | C. Huang, S. Wen, M. Li, F. Wen, X. Yang, An empirical evaluation of the influential nodes for stock market network: Chinese a shares case, Financ. Res. Lett., (2020), 101517. |
[9] | X. Yang, S. Wen, Z. Liu, C. Li, Dynamic properties of foreign exchange complex network, Mathematics, 7 (2019), 832. |
[10] | C. Huang, H. Yang, J. Cao, Weighted pseudo almost periodicity of multi-proportional delayed shunting inhibitory cellular neural networks with D operator, Discrete Contin. Dyn. Syst. Ser. S, 2020. |
[11] | C. Huang, X. Yang, J. Cao, Stability analysis of Nicholson's blowflies equation with two different delays, Math. Comput. Simulation, 171 (2020), 201-206. |
[12] | M. Iswarya, R. Raja, G. Rajchakit, J. Cao, J. Alzabut, C, Huang, Existence, uniqueness and exponential stability of periodic solution for discrete-time delayed BAM neural networks based on coincidence degree theory and graph theoretic method, Mathematics, 7 (2019), 1055. |
[13] | C. Huang, Z. Yang, T. Yi, X. Zou, On the basins of attraction for a class of delay differential equations with non-monotone bistable nonlinearities, J. Differential Equations, 256 (2014), 2101- 2114. |
[14] | C. Huang, J. Cao, F. Wen, X. Yang, Stability analysis of SIR model with distributed delay on complex networks, PLoS One, 11 (2016), e0158813. |
[15] | M. Mackey, L. Glass, Oscillation and chaos in physiological control systems, Science, 197 (1977), 287-289. |
[16] | C. Huang, X. Long, J. Cao, Stability of anti-periodic recurrent neural networks with multiproportional delays, Math. Meth. Appl. Sci., 43 (2020), 6093-6102. |
[17] | L. Berezansky, E. Braverman, Mackey-Glass equation with variable coefficients, Appl. Math. Comput., 51 (2006), 1-16. |
[18] | L. Berezansky, E. Braverman, L. Idels, Mackey-Glass model of hematopoiesis with non-monotone feedback: Stability, oscillation and control, Appl. Math. Comput., 219 (2013), 6268-6283. |
[19] | W. Wang, Global exponential stability for the Mackey-Glass model of respiratory dynamics with a control term, Math. Meth. Appl. Sci., 39 (2016), 606-613. |
[20] | L. Berezansky, E. Braverman, A note on stability of Mackey-Glass equations with two delays, J. Math. Anal. Appl., 450 (2017), 1208-1228. |
[21] | C. Huang, J. Wang, L. Huang, New results on asymptotically almost periodicity of delayed Nicholson-type system involving patch structure, Electron. J. Differ. Equ., 2020 (2020), 1-17. |
[22] | J. Alzabut, Y. Bolat, T. Abdeljawad, Almost periodic dynamics of a discrete Nicholson's blowflies model involving a linear harvesting term, Adv. Differ. Equ., 2012 (2012), 158. |
[23] | S. Saker, J. Alzabut, Periodic solutions, global attractivity and oscillation of an impulsive delay host-macroparasite model, Math. Comput. Model, 45 (2007), 531-543. |
[24] | S. Saker, J. Alzabut, On the impulsive delay hematopoiesis model with periodic coefficients, Rocky Mountain J. Math., 39 (2009), 1657-1688. |
[25] | J. Alzabut, J. Nieto, G. Stamov, Existence and exponential stability of positive almost periodic solutions s for a model of hematopoiesis, Bound. Value Probl., (2009), 127510. |
[26] | J. Alzabut, Almost periodic solutions for an impulsive delay Nicholson's blowflies model, J. Comput. Appl. Math., 234 (2010), 233-239. |
[27] | Y. Xu, Q. Cao, X. Guo, Stability on a patch structure Nicholson's blowflies systeminvolving distinctive delays, Appl. Math. Lett., 105 (2020), 106340. |
[28] | C. Huang, Y. Qiao, L. Huang, R. Agarwal, Dynamical behaviors of a food-chain model with stage structure and time delays, Adv. Difference Equ., 2018 (2018), 186. |
[29] | L. Duan, X. Fang, C. Huang, Global exponential convergence in a delayed almost periodic Nicholson's blowflies model with discontinuous harvesting, Math. Meth. Appl.Sci., 41 (2018), 1954-1965. |
[30] | H. Smith, An Introduction to Delay Differential Equations with Applications to the Life Sciences, Springer New York, 2011. |
[31] | H. A. El-Morshedy, A. Ruiz-Herrera, Global convergence to equilibria in non-monotone delay differential equations, Proc. Amer. Math. Soc., 147 (2019), 2095-2105. |
[32] | I. Győri, F. Hartung, N. A. Mohamady, Permanence in a class of delay differential equations with mixed monotonicity, Electron. J. Qual. Theory Differ. Equ., 53 (2018), 1-21. |
[33] | L. Berezansky, E. Braverman, L. Idels, The Mackey-Glass model of respiratory dynamics: Review and new results, Nonlinear Anal., 75 (2012), 6034-6052. |
[34] | L. Berezansky, E. Braverman, L. Idels, Mackey-Glass model of hematopoiesis with monotone feedback revisited, Appl. Math. Comput., 219 (2013), 4892-4907. |
1. | Qian Cao, Xin Long, New convergence on inertial neural networks with time-varying delays and continuously distributed delays, 2020, 5, 2473-6988, 5955, 10.3934/math.2020381 | |
2. | Xin Long, Novel stability criteria on a patch structure Nicholson’s blowflies model with multiple pairs of time-varying delays, 2020, 5, 2473-6988, 7387, 10.3934/math.2020473 | |
3. | Chuangxia Huang, Xian Zhao, Jinde Cao, Fuad E Alsaadi, Global dynamics of neoclassical growth model with multiple pairs of variable delays, 2020, 33, 0951-7715, 6819, 10.1088/1361-6544/abab4e | |
4. | Qian Cao, Guoqiu Wang, New findings on global exponential stability of inertial neural networks with both time-varying and distributed delays, 2021, 0952-813X, 1, 10.1080/0952813X.2021.1883744 | |
5. | Xianhui Zhang, Convergence analysis of a patch structure Nicholson’s blowflies system involving an oscillating death rate, 2021, 0952-813X, 1, 10.1080/0952813X.2021.1908433 | |
6. | Hui Zhou, Wen Wang, Liu Yang, Stage-structured hematopoiesis model with delays in an almost periodic environment, 2021, 08939659, 107336, 10.1016/j.aml.2021.107336 | |
7. | Zao Fu, Xing He, Ping Liu, Ali Palizban, Wengjing Liao, Distributed Neural Network and Particle Swarm Optimization for Micro-grid Adaptive Power Allocation, 2022, 54, 1370-4621, 3215, 10.1007/s11063-022-10760-6 | |
8. | Wenke Wang, Le Li, Xuejun Yi, Chuangxia Huang, Convergence on Population Dynamics and High-Dimensional Haddock Conjecture, 2021, 13, 2073-8994, 2252, 10.3390/sym13122252 | |
9. | Qian Cao, Attractivity analysis on a neoclassical growth system incorporating patch structure and multiple pairs of time-varying delays, 2021, 14173875, 1, 10.14232/ejqtde.2021.1.76 | |
10. | Xiaoyuan Chang, Junping Shi, Bistable and oscillatory dynamics of Nicholson's blowflies equation with Allee effect, 2022, 27, 1531-3492, 4551, 10.3934/dcdsb.2021242 | |
11. | Wenjie Hu, Quanxin Zhu, Spatial-temporal dynamics of a non-monotone reaction-diffusion Hopfield’s neural network model with delays, 2022, 34, 0941-0643, 11199, 10.1007/s00521-022-07036-4 | |
12. | Jian Zhang, Ancheng Chang, Gang Yang, Periodicity on Neutral-Type Inertial Neural Networks Incorporating Multiple Delays, 2021, 13, 2073-8994, 2231, 10.3390/sym13112231 | |
13. | Wentao Wang, Wei Chen, Mean-square exponential input-to-state stability of stochastic inertial neural networks, 2021, 2021, 1687-1847, 10.1186/s13662-021-03586-4 | |
14. | Xiaodan Ding, Global attractivity of Nicholson’s blowflies system with patch structure and multiple pairs of distinct time-varying delays, 2023, 16, 1793-5245, 10.1142/S1793524522500814 | |
15. | Ramazan Yazgan, Salsabil Hajjaji, Farouk Chérif, Weighted Pseudo Almost-Automorphic Solutions of Quaternion-Valued RNNs With Mixed Delays, 2023, 55, 1370-4621, 423, 10.1007/s11063-022-10890-x | |
16. | Chuangxia Huang, Bingwen Liu, Chaofan Qian, Jinde Cao, Stability on positive pseudo almost periodic solutions of HPDCNNs incorporating D operator, 2021, 190, 03784754, 1150, 10.1016/j.matcom.2021.06.027 | |
17. | Rajendiran Eswari, Jehad Alzabut, Mohammad Esmael Samei, Hui Zhou, On periodic solutions of a discrete Nicholson’s dual system with density-dependent mortality and harvesting terms, 2021, 2021, 1687-1847, 10.1186/s13662-021-03521-7 | |
18. | Yang Gao, Global stability for a new predator–prey model with cross-dispersal among patches based on graph theory, 2021, 2021, 1687-1847, 10.1186/s13662-021-03645-w | |
19. | Lilun Zhang, Le Li, Chuangxia Huang, Positive stability analysis of pseudo almost periodic solutions for HDCNNs accompanying $ D $ operator, 2022, 15, 1937-1632, 1651, 10.3934/dcdss.2021160 | |
20. | Qian Cao, Positive almost periodicity on a n-species food chain system incorporating time-varying delays and polluted environments, 2022, 0952-813X, 1, 10.1080/0952813X.2022.2080279 | |
21. | Liang Zhou, Huawei Wang, Multihorizons transfer strategy for continuous online prediction of time‐series data in complex systems, 2022, 37, 0884-8173, 7706, 10.1002/int.22900 | |
22. | Chuangxia Huang, Xiaodan Ding, Existence of traveling wave fronts for a diffusive Mackey–Glass model with two delays, 2024, 76, 14681218, 104024, 10.1016/j.nonrwa.2023.104024 | |
23. | V. Alekseev, Two-cluster synchronization in a fully coupled network of Mackey–Glass generators, 2024, 12, 26668181, 100930, 10.1016/j.padiff.2024.100930 | |
24. | Adrian Gomez, Nolbert Morales, Manuel Zamora, Non-Trivial Periodic Solutions for a Class of Second Order Differential Equations with Large Delay, 2023, 188, 0167-8019, 10.1007/s10440-023-00613-2 | |
25. | Divya Agrawal, Soniya Dhama, Marko Kostić, Syed Abbas, Periodicity, stability, and synchronization of solutions of hybrid coupled dynamic equations with multiple delays, 2024, 47, 0170-4214, 7616, 10.1002/mma.9993 |