Abstract

By using fixed-point theorem and under suitable conditions, we investigate the existence and multiplicity positive solutions to the following systems: 𝑢(𝑡)+𝑎𝑢(𝑡)+𝑏𝑣(𝑡)+𝜆1(𝑡)𝑓(𝑢(𝑡),𝑣(𝑡))=0,𝑡[0,1],𝑣(𝑡)+𝑐𝑢(𝑡)+𝑑𝑣(𝑡)+𝜇2(𝑡)𝑔(𝑢(𝑡),𝑣(𝑡))=0,𝑡[0,1],𝑢(0)=𝑢(1)=0,𝑣(0)=𝑣(1)=0, where 𝑎,𝑏,𝑐,𝑑 are four positive constants and 𝜆>0, 𝜇>0, 𝑓(𝑢,𝑣),𝑔(𝑢,𝑣)𝐶(𝑅+×𝑅+,𝑅+) and 1,2𝐶([0,1],𝑅+). We derive two explicit intervals of 𝜆 and 𝜇, such that the existence and multiplicity of positive solutions for the systems is guaranteed.

1. Introduction

In this paper, we are concerned with the existence and multiplicity of positive solutions to the following boundary value problem of second-order nonlinear differential systems with two parameters: 𝑢(𝑡)+𝑎𝑢(𝑡)+𝑏𝑣(𝑡)+𝜆1[],𝑣(𝑡)𝑓(𝑢(𝑡),𝑣(𝑡))=0,𝑡0,1(𝑡)+𝑐𝑢(𝑡)+𝑑𝑣(𝑡)+𝜇2[],𝑣(𝑡)𝑔(𝑢(𝑡),𝑣(𝑡))=0,𝑡0,1𝑢(0)=𝑢(1)=0,(0)=𝑣(1)=0,(1.1) where 𝑎,𝑏,𝑐,𝑑 are four positive constants.

In addition, we assume the following conditions throughout this paper:(𝐻1)1(𝑡),2(𝑡)𝐶([0,1],𝑅+) does not vanish identically on any subinterval of [0,1];(𝐻2)0<𝑎<𝜋2,0<𝑑<𝜋2,𝑐>0,𝑏>0; (𝐻3)𝑓,𝑔𝐶(𝑅+×𝑅+,𝑅+).

The boundary value problem of ordinary differential systems has attracted much attention, see [16] and the references therein. Recently, Fink and Gatica [4] and Ma [7] have studied the existence of positive solutions of the following systems: 𝑥[],𝑦(𝑡)+𝜆𝑎(𝑡)𝑓(𝑥(𝑡),𝑦(𝑡))=0,𝑡0,1[],(𝑡)+𝜇𝑏(𝑡)𝑔(𝑥(𝑡),𝑦(𝑡))=0,𝑡0,1𝑥(0)=𝑥(1)=𝑦(0)=𝑦(1)=0.(1.2) In [4], a multiplicity result has been established when 𝑓(0,0)>0,𝑔(0,0)>0. In [7], a multiplicity result has been given for the more general case 𝑓(0,0)0,𝑔(0,0)0.

Also, by using Krasnoselskill fixed-point theorem, Ru and An [8] considered the existence of positive solutions of the following systems: ()𝑝𝑢(2𝑝)[],=𝜆𝑎(𝑡)𝑓(𝑢(𝑡),𝑣(𝑡)),𝑡0,1()𝑞𝑣(2𝑞)[],𝑢=𝜇𝑏(𝑡)𝑔(𝑢(𝑡),𝑣(𝑡)),𝑡0,1(2𝑖)(0)=𝑢(2𝑖)𝑣(1)=0,0𝑖𝑝1,(2𝑗)(0)=𝑣(2𝑗)(1)=0,0𝑗𝑞1,(1.3) where 𝜆>,𝜇>0, 𝑝,𝑞𝑁.𝑓,𝑔[0,)×[0,)[0,).

More recently, Dalbono and Mckenna [5] proved the existence and multiplicity of solutions to a class of asymmetric weakly coupled systems as follows: 𝑢1𝑢(𝑡)2(+𝑏𝑡)1𝜀𝜀𝑏2𝑢+1𝑢(𝑡)+2(=,𝑢𝑡)sin(𝑡)sin(𝑡)1(0)=𝑢2(0)=0=𝑢1(𝜋)=𝑢2(𝜋),(1.4) where 𝜀 is suitably small and the positive numbers 𝑏1, 𝑏2 satisfy 2<𝑏1<(+1)2,𝑘2<𝑏2<(𝑘+1)2forsome,𝑘𝑁.(1.5) Applying a classical change of variables, the authors transformed the initial problem into an equivalent problem whose solutions can be characterized by their nodal properties. Meanwhile, in [5], there are two open questions. (1)Can one replace the near-diagonal matrix with something more general and use information on the eigenvalues of matrix?(2)Can one replace the near-diagonal matrix with something more general and use information on the eigenvalues of matrix?

Inspired by the above works and the two open questions, we consider the existence and multiplicity of positive solutions to (1.1). The paper is organized as follows. In Section 2, we state some preliminaries. In Sections 3 and 4, we prove the existence and multiplicity results of (1.1).

2. Preliminaries

In order to prove our results, we state the well-known fixed-point theorem [9]:

Lemma 2.1. Let 𝐸 be a Banach space and let 𝐾𝐸 be a cone in 𝐸. Assume Ω1, Ω2 are two open subsets of 𝐸 with 0Ω1, Ω1Ω2, and let 𝑇𝐾(Ω2Ω1)𝐾 be a completely continuous operator such that either(i)𝑇𝑢𝑢,𝑢𝐾𝜕Ω1 and 𝑇𝑢𝑢,𝑢𝐾𝜕Ω2; or(ii)𝑇𝑢𝑢,𝑢𝐾𝜕Ω1 and 𝑇𝑢𝑢,𝑢𝐾𝜕Ω2.Then, 𝑇 has a fixed point in 𝐾(Ω2Ω1).
To be convenient, we introduce the following notations: 𝑓0=lim𝑢+𝑣0𝑓(𝑢,𝑣)𝑢+𝑣,𝑓=lim𝑢+𝑣𝑓(𝑢,𝑣),𝑔𝑢+𝑣0=lim𝑢+𝑣0𝑔(𝑢,𝑣)𝑢+𝑣,𝑔=lim𝑢+𝑣𝑔(𝑢,𝑣),𝑢+𝑣(2.1) and suppose that 𝑓0,𝑔0,𝑓,𝑔[0,+].
Let 𝜔1=𝑎,𝜔2=𝑑,𝐻(𝑡)𝐶[0,1], 𝐺𝑖(𝑡,𝑠),𝑖=1,2 be 𝐺𝑟𝑒𝑒𝑛𝑠 function of the corresponding to linear boundary value problem: 𝑢(𝑡)𝜔2𝑖𝑢(𝑡)=𝐻(𝑡),𝑢(0)=𝑢(1)=0.(2.2) Then, the solution of (2.2) is given by 𝑢(𝑡)=10𝐺𝑖(𝑡,𝑠)𝐻(𝑠)𝑑𝑠.(2.3) It is well known that 𝐺𝑖(𝑡,𝑠) can be expressed by 𝐺𝑖(𝑡,𝑠)=sin𝜔𝑖𝑡sin𝜔𝑖(1𝑠)𝜔𝑖sin𝜔𝑖0𝑡𝑠1,sin𝜔𝑖𝑠sin𝜔𝑖(1𝑡)𝜔𝑖sin𝜔𝑖0𝑠𝑡1.(2.4) In addition, it can be easily to be checked that 𝑀𝑖=max[]𝑡0,1𝐺𝑖(𝑡,𝑡)>0,𝑚𝑖=min[]𝑡1/4,3/4𝐺𝑖(𝑡,𝑡)>0,𝑖=1,2.(2.5)

See [6], we have the following lemma.

Lemma 2.2. 𝐺𝑖(𝑡,𝑠) has the following properties:(i)𝐺𝑖(𝑡,𝑠)>0,forall𝑡,𝑠(0,1), (ii)𝐺𝑖(𝑡,𝑠)𝐶𝑖𝐺𝑖(𝑠,𝑠),forall𝑡,𝑠[0,1],𝐶𝑖=1/sin𝜔𝑖, (iii)𝐺𝑖(𝑡,𝑠)𝛿𝑖𝐺𝑖(𝑡,𝑡)𝐺𝑖(𝑠,𝑠),forall𝑡,𝑠[0,1],𝛿𝑖=𝜔𝑖sin𝜔𝑖,𝑖=1,2. It is obvious that problem (1.1) is equivalent to the equation: (𝑢(𝑡),𝑣(𝑡))=10𝐺1(𝑡,𝑠)𝜆1(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑s,10𝐺2(𝑡,𝑠)𝜇2,(𝑠)𝑔(𝑢(𝑠),𝑣(𝑠))+𝑐𝑢(𝑠)𝑑𝑠(2.6) and consequently it is equivalent to the fixed-point problem: (𝑢,𝑣)=𝐀(𝑢,𝑣)(2.7) with 𝐀𝐶[0,1]×𝐶[0,1]𝐶[0,1]×𝐶[0,1] given by =𝐀(𝑢,𝑣)10𝐺1(𝑡,𝑠)𝜆1(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠,10𝐺2(𝑡,𝑠)𝜇2.(𝑠)𝑔(𝑢(𝑠),𝑣(𝑠))+𝑐𝑢(𝑠)𝑑𝑠(2.8) For convenience, denote 𝐀𝜆(𝑢,𝑣)=10𝐺1(𝑡,𝑠)𝜆1𝐀(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠,𝜇(𝑢,𝑣)=10𝐺2(𝑡,𝑠)𝜇2(𝑠)𝑔(𝑢(𝑠),𝑣(𝑠))+𝑐𝑢(𝑠)𝑑𝑠.(2.9) It is obvious that 𝐀 is completely continuous. Let 𝜎𝑖=𝛿𝑖𝑚𝑖/𝐶𝑖=(sin𝜔𝑖/4)(sin3𝜔𝑖/4)(sin𝜔𝑖), 𝑖=1,2,𝜎=min{𝜎1,𝜎2} and define a cone in 𝐶[0,1]×𝐶[0,1] by [][]𝐊=(𝑢,𝑣)𝐶0,1×𝐶0,1𝑢(𝑡)0,𝑣(𝑡)0,min1/4𝑡3/4(𝑢(𝑡)+𝑣(𝑡))𝜎(𝑢,𝑣),(2.10) where (𝑢,𝑣)=𝑢+𝑣=sup𝑡[0,1]𝑢(𝑡)+sup𝑡[0,1]𝑣(𝑡).

Lemma 2.3. 𝐴(𝐾)𝐾. Proof. For any (𝑡,𝑠)[1/4,3/4]×[0,1], by Lemma 2.2, we have 𝐴𝜆(𝑢,𝑣)(𝑡)=10𝐺1(𝑡,𝑠)𝜆1(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠10𝐶1𝐺1(𝑠,𝑠)𝜆1𝐴(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠,𝜆(𝑢,𝑣)10𝐶1𝐺1(𝑠,𝑠)𝜆1𝐴(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠,𝜆(𝑢,𝑣)(𝑡)=10𝐺1(𝑡,𝑠)𝜆1(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠𝛿1𝐺1(𝑡,𝑡)10𝐺1(𝑠,𝑠)𝜆1𝛿(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠1𝐺1(𝑡,𝑡)𝐶1𝐴𝜆(.𝑢,𝑣)(2.11) Similarly, for any (𝑡,𝑠)[1/4,3/4]×[0,1], we have 𝐴𝜇𝛿(𝑢,𝑣)(𝑡)2𝐺2(𝑡,𝑡)𝐶2𝐴𝜇.(𝑢,𝑣)(2.12) Then, for any (𝑡,𝑠)[1/4,3/4]×[0,1], we have 𝐴𝜆(𝑢,𝑣)(𝑡)+𝐴𝜇(𝑢,𝑣)(𝑡)=10𝐺1(𝑡,𝑠)𝜆1𝑓(𝑢,𝑣)+𝑏𝑣𝑑𝑠+10𝐺2(𝑡,𝑠)𝜇2𝛿𝑔(𝑢,𝑣)+𝑐𝑢𝑑𝑠1𝐺1(𝑡,𝑡)𝐶1𝐴𝜆+𝛿(𝑢,𝑣)2𝐺2(𝑡,𝑡)𝐶2𝐴𝜇(𝑢,𝑣)𝜎𝐴(𝑢,𝑣).(2.13) Thus, min1/4𝑡3/4𝐴𝜆(𝑢,𝑣)+𝐴𝜇(𝑢,𝑣)𝜎𝐴(𝑢,𝑣). Therefore, 𝐴(𝐾)𝐾.

3. Existence Results

We assume the following:(𝐻4)0<𝑏<2𝜔1cos2(𝜔1/2),0<𝑐<2𝜔2cos2(𝜔2/2); (𝐻5)𝐴𝑖=10𝑖(𝑠)𝑑𝑠,𝐵𝑖=3/41/4𝑖(𝑠)𝑑𝑠>0,𝑖=1,2; (𝐻6)𝑃𝑖=𝐶𝑖𝑀𝑖=(sec2𝜔𝑖/2)/4𝜔𝑖,𝑄𝑖=𝛿𝑖𝑚2𝑖=sin(𝜔𝑖/4)sin(3𝜔𝑖/4),𝑖=1,2.

Theorem 3.1. Assume (𝐻1)(𝐻6) hold. Then one has the following:
(1)If 0<𝑓0,𝑓,𝑔0,𝑔<,2𝑃1𝐴1𝑓0<𝑄1𝜎𝐵1𝑓(12𝑃1𝑏), then for each 𝜆(1/𝑄1𝜎𝐵1𝑓,12𝑃1𝑏/2𝑃1𝐴1𝑓0) and 𝜇(0,(12𝑃2𝑐)/2𝑃2𝐴2𝑔0), (1.1) has at least one positive solution. (2)If 0<𝑓0,𝑓,𝑔0,𝑔<,2𝑃2𝐴2𝑔0<𝑄2𝜎𝐵2𝑔(12𝑃2𝑐), then for each 𝜆(0,(12𝑃1𝑏)/2𝑃1𝐴1𝑓0) and 𝜇(1/𝑄2𝜎𝐵2𝑔,(12𝑃2𝑐)/2𝑃2𝐴2𝑔0), (1.1) has at least one positive solution. (3)If 𝑓0=0,𝑓=, then for each 𝜆(0,) and 𝜇(0,(12𝑃2𝑐)/2𝑃2𝐴2𝑔0), (1.1) has at least one positive solution.(4)If 𝑔0=0,𝑔=, then for each 𝜆(0,(12𝑃1𝑏)/2𝑃1𝐴1𝑓0) and 𝜇(0,), (1.1) has at least one positive solution.(5)If 𝑓0=0,𝑓= and 𝑔0=0,𝑔=, then for each 𝜆(0,) and 𝜇(0,), (1.1) has at least one positive solution.(6)If 𝑓=,0<𝑓0< or 𝑔=,0<𝑔0<, then for each 𝜆(0,(12𝑃1𝑏)/2𝑃1𝐴1𝑓0) and 𝜇(0,(12𝑃2𝑐)/2𝑃2𝐴2𝑔0), (1.1) has at least one positive solution. (7)If 𝑓0=0,0<𝑓< and 𝑔0=0,0<𝑔<, then for each 𝜆(1/𝑄1𝜎𝐵1𝑓,),𝜇(0,) or 𝜆(0,),𝜇((1/𝑄2𝜎𝐵2𝑔)),(1.1) has at least one positive solution.

Proof. We only prove case (1.1). The other cases can be proved similarly. In order to apply the Lemma 2.1, we construct the sets Ω1,Ω2.
Let 𝜆(1/𝑄1𝜎𝐵1𝑓,(12𝑃1𝑏)/2𝑃1𝐴1𝑓0),𝜇(0,(12𝑃2𝑐)/2𝑃2𝐴2𝑔0), and we choose 𝜀>0 such that 1𝜎𝑄1𝐵1𝑓𝜀𝜆12𝑃1𝑏2𝑃1𝐴1𝑓0+𝜀,0<𝜇12𝑃2𝑐2𝑃2𝐴2𝑔0+𝜀.(3.1)
By the definition of 𝑓0 and 𝑔0, there exists 𝑅1>0, such that 𝑓𝑓(𝑢,𝑣)0𝑔+𝜖(𝑢+𝑣),𝑔(𝑢,𝑣)0+𝜖(𝑢+𝑣),for𝑢+𝑣0,𝑅1.(3.2) Choosing (𝑢,𝑣)𝐾 with (𝑢,𝑣)=𝑅1, we have 𝐴𝜆(𝑢,𝑣)(𝑡)=10𝐺1(𝑡,𝑠)𝜆1(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠𝐶1𝑀110𝜆1𝑓(𝑠)0+𝜀(𝑢(𝑠)+𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠𝐶1𝑀1(𝑢,𝑣)10𝜆1𝑓(𝑠)0+𝜀+𝑏𝑑𝑠=𝑃1(𝑢,𝑣)𝜆𝐴1𝑓01+𝜀+𝑏2(𝑢,𝑣),(3.3) namely, 𝐴𝜆(𝑢,𝑣)(𝑡)(1/2)(𝑢,𝑣). In the same way, we also have 𝐴𝜇1(𝑢,𝑣)(𝑡)2(𝑢,𝑣).(3.4) then 𝐴𝜆(𝑢,𝑣)(𝑡)(𝑢,𝑣)(𝑡). Thus, if we set Ω1={(𝑢,𝑣)𝐾(𝑢,𝑣)<𝑅1}, then 𝐴(𝑢,𝑣)(𝑡)(𝑢,𝑣),(𝑢,𝑣)𝐾𝜕Ω1.(3.5) On the other hand, by the definition of 𝑓, there exists 𝑅2>0, such that 𝑓(𝑢,𝑣)(𝑓𝜖)(𝑢+𝑣), for 𝑢+𝑣[𝑅2,+). Let 𝑅2=max{2𝑅1,𝜎1𝑅2} and Ω2={(𝑢,𝑣)𝐾(𝑢,𝑣)<𝑅2}. If we choose (𝑢,𝑣)𝐾 with (𝑢,𝑣)=𝑅2, such that min1/4𝑡3/4(𝑢(𝑡)+𝑣(𝑡))𝜎(𝑢,𝑣)𝑅2, then we have 𝐴𝜆1(𝑢,𝑣)4=10𝐺114,𝑠𝜆1(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠𝛿1𝐺114,143/41/4𝐺1(𝑠,𝑠)𝜆1(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠𝛿1𝑚213/41/4𝜆1𝑓(𝑠)𝜀(𝑢(𝑠)+𝑣(𝑠))𝑑𝑠𝑄1𝜎𝐵1𝜆𝑓𝜀(𝑢,𝑣)(𝑢,𝑣).(3.6) Hence 𝐴𝐴(𝑢,𝑣)(𝑡)𝜆(𝑢,𝑣)(𝑡)(𝑢,𝑣),(𝑢,𝑣)𝐾𝜕Ω2.(3.7) Therefore, it follows from (3.5) to (3.7) and Lemma 2.2, 𝐴 has a fixed point (𝑢,𝑣)𝐾(Ω2Ω1), which is a positive solution of (1.1).

Similarly, we also have the following results.

Theorem 3.2. Assume (𝐻1)(𝐻6) hold. Then one has the following:(1)If 0<𝑓0,𝑓,𝑔0,𝑔<,2𝑃1𝐴1𝑓<𝑄1𝜎𝐵1𝑓0(12𝑃1𝑏), then for each 𝜆(1/𝑄1𝜎𝐵1𝑓0,(12𝑃1𝑏)/2𝑃1𝐴1𝑓) and 𝜇(0,(12𝑃2𝑐)/2𝑃2𝐴2𝑔), (1.1) has at least one positive solution.(2)If 0<𝑓0,𝑓,𝑔0,𝑔<,2𝑃2𝐴2𝑔<𝑄2𝜎𝐵2𝑔0(12𝑃2𝑐), then for each 𝜆(0,(12𝑃1𝑏)/2𝑃1𝐴1𝑓) and 𝜇(1/𝑄2𝜎𝐵2𝑔0,(12𝑃2𝑐)/2𝑃2𝐴2𝑔), (1.1) has at least one positive solution.(3)If 𝑓0=,𝑓=0, then for each 𝜆(0,) and 𝜇(0,(12𝑃2𝑐)/2𝑃2𝐴2𝑔0), (1.1) has at least one positive solution.(4)If 𝑔0=,𝑔=0, then for each 𝜆(0,(12𝑃1𝑏)/2𝑃1𝐴1𝑓0) and 𝜇(0,), (1.1) has at least one positive solution.(5)If 𝑓0=,𝑓=0 and 𝑔0=,𝑔=0, then for each 𝜆(0,) and 𝜇(0,), (1.1) has at least one positive solution.(6)If 𝑓0=,0<𝑓< or 𝑔0=,0<𝑔< then for each 𝜆(0,(12𝑃1𝑏)/2𝑃1𝐴1𝑓) and 𝜇(0,(12𝑃2𝑐)/2𝑃2𝐴2𝑔), (1.1) has at least one positive solution.(7)If 𝑓=0,0<𝑓0< and 𝑔=0,0<𝑔0< then for each 𝜆(1/𝑄1𝜎𝐵1𝑓0,),𝜇(0,) or 𝜆(0,),𝜇(1/𝑄2𝜎𝐵2𝑔0), (1.1) has at least one positive solution.

Proof. We only prove case (1). The other cases can be proved similarly.
Let 𝜆(1/𝑄1𝜎𝐵1𝑓0,(12𝑃1𝑏)/2𝑃1𝐴1𝑓),𝜇(0,(12𝑃2𝑐)/2𝑃2𝐴2𝑔), and we choose 𝜀>0 such that 1𝜎𝑄1𝐵1𝑓0𝜀𝜆12𝑃1𝑏2𝑃1𝐴1𝑓+𝜀,0<𝜇12𝑃2𝑐2𝑃2𝐴2𝑔+𝜀,(3.8)
by the definition of 𝑓0, there exists 𝑅1>0, such that 𝑓𝑓(u,𝑣)0𝜖(𝑢+𝑣),for𝑢+𝑣0,𝑅1.(3.9) Choosing (𝑢,𝑣)𝐾 with (𝑢,𝑣)=𝑅1, such that min1/4𝑡3/4(𝑢(𝑡)+𝑣(𝑡))𝜎(𝑢,𝑣), then we have 𝐴𝜆1(𝑢,𝑣)4=10𝐺114,𝑠𝜆1(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠𝛿1𝐺114,143/41/4𝐺1(𝑠,𝑠)𝜆1(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠𝛿1𝑚213/41/4𝜆1𝑓(𝑠)0𝜀(𝑢(𝑠)+𝑣(𝑠))𝑑𝑠𝑄1𝜎𝐵1𝜆𝑓0𝜀(𝑢,𝑣)(𝑢,𝑣).(3.10) So, if we set Ω1={(𝑢,𝑣)𝐾(𝑢,𝑣)<𝑅1}, then 𝐴(𝑢,𝑣)(𝑡)(𝑢,𝑣),(𝑢,𝑣)𝐾𝜕Ω1.(3.11) On the other hand, by the definition of 𝑓 and 𝑔, there exists 𝑅2>2𝑅1, such that 𝑓𝑓(𝑢,𝑣)𝑔+𝜖(𝑢+𝑣),𝑔(𝑢,𝑣)𝑅+𝜖(𝑢+𝑣)for𝑢+𝑣2,+.(3.12) Choosing (𝑢,𝑣)𝐾 with (𝑢,𝑣)=𝑅2, we have 𝐴𝜆(𝑢,𝑣)(𝑡)=10𝐺1(𝑡,𝑠)𝜆1(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠10𝐶1𝐺1(𝑠,𝑠)𝜆1(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠𝐶1𝑀110𝜆1𝑓(𝑠)+𝜀(𝑢(𝑠)+𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠𝑃1𝜆𝐴1𝑓1+𝜀+𝑏(𝑢,𝑣)2(𝑢,𝑣).(3.13) In the same way, we also have 𝐴𝜇1(𝑢,𝑣)(𝑡)2(𝑢,𝑣).(3.14) Hence, if we set Ω2={(𝑢,𝑣)𝐾(𝑢,𝑣)<𝑅2}, then 𝐴(𝑢,𝑣)(𝑡)(𝑢,𝑣)(𝑡),(𝑢,𝑣)𝐾𝜕Ω2.(3.15) Therefore, it follows from (3.11) to (3.15) and Lemma 2.2 that 𝐴 has a fixed point (𝑢,𝑣)𝐾(Ω2Ω1), which is a positive solution of (1.1).

4. Multiplicity Results

Theorem 4.1. Assume (𝐻1)(𝐻6) hold. In addition, assume that there exist three constants 𝑟,𝑀,𝑁, where 𝑁 is sufficient small with 2𝑃1𝑁𝐴1𝑄1𝐵1𝑀(12𝑃1𝑏),2𝑃2𝑁𝐴2𝑄2𝐵2𝑀(12𝑃2𝑐) such that(i)𝑓0=𝑓=0,𝑔0=𝑔=0, (ii)𝑓(𝑢,𝑣)𝑀𝑟𝑜𝑟𝑔(𝑢,𝑣)𝑀𝑟,  for 𝜎𝑟(𝑢,𝑣)𝑟.Then, for any 𝜆(1/𝑄1𝐵1𝑀,(12𝑃1𝑏)/2𝑃1𝐴1𝑁), 𝜇(0,(12𝑃2𝑐)/2𝑃2𝐴2𝑁), or 𝜆(0,(12𝑃1𝑏)/2𝑃1𝐴1𝑁), 𝜇(1/𝑄2𝐵2𝑀,(12𝑃2𝑐)/2𝑃2𝐴2𝑁), the problem (1.1) has at least two positive solutions.

Proof. We only prove the case of 𝜆(1/𝑄1𝜎𝐵1𝑀,(12𝑃1𝑏)/2𝑃1𝐴1𝑁), 𝜇(0,(12𝑃2𝑐/2𝑃2𝐴2𝑁)), The other case is similar.
Step 1. Since 𝑓0=𝑔0=0, there exists 𝑟1(0,𝑟) such that 𝑓(𝑢,𝑣)𝑁(𝑢+𝑣),𝑔(𝑢,𝑣)𝑁(𝑢+𝑣),for0<𝑢+𝑣<𝑟1.(4.1) Set Ω1={(𝑢,𝑣)𝐾(𝑢,𝑣)<𝑟1},forall(𝑢,𝑣)𝐾𝜕Ω1, then we have 𝐴𝜆(𝑢,𝑣)(𝑡)=10𝐺1(𝑡,𝑠)𝜆1(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠10𝐶1𝑀1𝜆1(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠𝐶1𝑀110𝜆1(𝑠)𝑁(𝑢(𝑠)+𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠𝑃1𝜆𝐴11𝑁+𝑏(𝑢,𝑣)2(𝑢,𝑣).(4.2) In the same way, we also have 𝐴𝜇1(𝑢,𝑣)(𝑡)2(𝑢,𝑣).(4.3) Hence, 𝐴(𝑢,𝑣)(𝑢,𝑣),(𝑢,𝑣)𝐾𝜕Ω1.(4.4)
Step 2. Since 𝑓=𝑔=0, there exists 𝑅>𝑟 such that 𝑓(𝑢,𝑣)𝑁(𝑢+𝑣),𝑔(𝑢,𝑣)𝑁(𝑢+𝑣),for𝑢+𝑣𝑅.(4.5) Similarly, set Ω2={(𝑢,𝑣)𝐾(𝑢,𝑣)<𝑟2}, then 𝐴(𝑢,𝑣)(𝑢,𝑣),(𝑢,𝑣)𝐾𝜕Ω2.(4.6)
Step 3. Let Ω3={(𝑢,𝑣)𝐾(𝑢,𝑣)<𝑟}, we can see that 𝐴𝜆1(𝑢,𝑣)4=10𝐺114,𝑠𝜆1(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠10𝛿1𝐺114,14𝐺1(𝑠,𝑠)𝜆1(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠𝛿1𝐺114,143/41/4𝐺1(𝑠,𝑠)𝜆1(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠𝑄13/41/4𝜆1(𝑠)(𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠))𝑑𝑠𝑄1𝜆𝐵1𝑀𝑟𝑟.(4.7) Then, 𝐴(𝑢,𝑣)(𝑢,𝑣),(𝑢,𝑣)𝐾𝜕Ω3.(4.8)
Consequently, by Lemma 2.1 and from (4.4)–(4.8), 𝐴 has two fixed points: (𝑢1,𝑣1)𝐾(Ω3Ω1) and (𝑢2,𝑣2)𝐾(Ω2Ω3), so (1.1) has at least two positive solutions satisfying 0<(𝑢1,𝑣1)<𝑟<(𝑢2,𝑣2).

Theorem 4.2. Assume (𝐻1)(𝐻6) hold. In addition, assume that there exist constants 𝑟,𝑀,𝑁, where 𝑁 is sufficient large with 2𝑃1𝑀𝐴1𝑄1𝜎𝐵1𝑁(12𝑃1𝑏),2𝑃2𝑀𝐴2𝑄2𝜎𝐵2𝑁(12𝑃2𝑐) such that(i)𝑓0=𝑓= or 𝑔0=𝑔=,(ii)𝑓(𝑢,𝑣)𝑀𝑟 or 𝑔(𝑢,𝑣)𝑀𝑟,  for 𝜎𝑟(𝑢,𝑣)𝑟,then, for any 𝜆(1/𝑄1𝜎𝐵1𝑁,(12𝑃1𝑏)/2𝑃1𝐴1𝑀), 𝜇(0,(12𝑃2𝑐)/2𝑃2𝐴2𝑀), or 𝜆(0,(12𝑃1𝑏)/2𝑃1𝐴1𝑀), 𝜇(1/𝑄2𝜎𝐵2𝑁,(12𝑃2𝑐)/2𝑃2𝐴2𝑀), the problem (1.1) has at least two positive solutions.

Proof. We only prove the case of 𝜆(1/𝑄1𝜎𝐵1𝑁,(12𝑃1𝑏)/2𝑃1𝐴1𝑀), 𝜇(0,(12𝑃2𝑐)/2𝑃2𝐴2𝑀), The other case is similar.
Step 1. Since 𝑓0=, there exists 𝑟1(0,𝑟) such that 𝑓(𝑢,𝑣)𝑁(𝑢+𝑣),for0<𝑢+𝑣𝑟1.(4.9) Set Ω1={(𝑢,𝑣)𝐾(𝑢,𝑣)<𝑟1}, then we have 𝐴𝜆1(𝑢,𝑣)4=10𝐺114,𝑠𝜆1(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠10𝛿1𝐺114,14𝐺1(𝑠,𝑠)𝜆1(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠𝛿1𝐺114,143/41/4𝐺1(𝑠,𝑠)𝜆1(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠𝑄13/41/4𝜆1(𝑠)(𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠))𝑑𝑠𝑄1𝜆𝐵1𝑁𝜎(𝑢,𝑣)𝑟.(4.10) Hence, 𝐴(𝑢,𝑣)(𝑢,𝑣),(𝑢,𝑣)𝐾𝜕Ω1.(4.11)
Step 2. Since 𝑓=, there exists 𝑅>𝑟 such that 𝑓(𝑢,𝑣)𝑁(𝑢+𝑣),for𝑢+𝑣𝑅.(4.12) Similarly, set Ω2={(𝑢,𝑣)𝐾(𝑢,𝑣)<𝑟2}, then 𝐴(𝑢,𝑣)(𝑢,𝑣),(𝑢,𝑣)𝐾𝜕Ω2.(4.13)
Step 3. Let Ω3={(𝑢,𝑣)𝐾(𝑢,𝑣)<𝑟},forall(𝑢,𝑣)𝐾𝜕Ω3, we can see that 𝐴𝜆(𝑢,𝑣)(𝑡)=10𝐺1(𝑡,𝑠)𝜆1(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠10𝐶1𝑀1𝜆1(𝑠)𝑓(𝑢(𝑠),𝑣(𝑠))+𝑏𝑣(𝑠)𝑑𝑠𝑃1𝜆𝐴11𝑀+𝑏(𝑢,𝑣)2.(𝑢,𝑣)(4.14) In the same way, we also have 𝐴𝜇1(𝑢,𝑣)(𝑡)2(𝑢,𝑣).(4.15) Hence, 𝐴(𝑢,𝑣)(𝑢,𝑣),(𝑢,𝑣)𝐾𝜕Ω3.(4.16)

Consequently, by Lemma 2.1 and from (4.11)–(4.16), 𝐴 has two fixed points: (𝑢1,𝑣1)𝐾(Ω3Ω1) and (𝑢2,𝑣2)𝐾(Ω2Ω3), so (1.1) has at least two positive solutions satisfying 0<(𝑢1,𝑣1)<𝑟<(𝑢2,𝑣2).