Under addressing global competition, manufacturing companies strive to produce better and cheaper products more quickly. For a complex production system, the design problem is intrinsically a daunting optimization task often involving multiple disciplines, nonlinear mathematical model, and computation-intensive processes during manufacturing process. Here is a reason to develop a high performance algorithm for finding an optimal solution to the engineering design and/or optimization problems. In this paper, a hybrid metaheuristic approach is proposed for solving engineering optimization problems. A genetic algorithm (GA), particle swarm optimization (PSO), and teaching and learning-based optimization (TLBO), called the GA-PSO-TLBO approach, is used and demonstrated for the proposed hybrid metaheuristic approach. Since each approach has its strengths and weaknesses, the GA-PSO-TLBO approach provides an optimal strategy that maintains the strengths as well as mitigates the weaknesses, as needed. The performance of the GA-PSO-TLBO approach is compared with those of conventional approaches such as single metaheuristic approaches (GA, PSO and TLBO) and hybrid metaheuristic approaches (GA-PSO and GA-TLBO) using various types of engineering optimization problems. An additional analysis for reinforcing the performance of the GA-PSO-TLBO approach was also carried out. Experimental results proved that the GA-PSO-TLBO approach outperforms conventional competing approaches and demonstrates high flexibility and efficiency.
Citation: YoungSu Yun, Mitsuo Gen, Tserengotov Nomin Erdene. Applying GA-PSO-TLBO approach to engineering optimization problems[J]. Mathematical Biosciences and Engineering, 2023, 20(1): 552-571. doi: 10.3934/mbe.2023025
[1] | Mustafa Mudhesh, Hasanen A. Hammad, Eskandar Ameer, Muhammad Arshad, Fahd Jarad . Novel results on fixed-point methodologies for hybrid contraction mappings in Mb-metric spaces with an application. AIMS Mathematics, 2023, 8(1): 1530-1549. doi: 10.3934/math.2023077 |
[2] | Muhammad Tariq, Muhammad Arshad, Mujahid Abbas, Eskandar Ameer, Saber Mansour, Hassen Aydi . A relation theoretic m-metric fixed point algorithm and related applications. AIMS Mathematics, 2023, 8(8): 19504-19525. doi: 10.3934/math.2023995 |
[3] | Huaping Huang, Bessem Samet . Two fixed point theorems in complete metric spaces. AIMS Mathematics, 2024, 9(11): 30612-30637. doi: 10.3934/math.20241478 |
[4] | Hanadi Zahed, Zhenhua Ma, Jamshaid Ahmad . On fixed point results in F-metric spaces with applications. AIMS Mathematics, 2023, 8(7): 16887-16905. doi: 10.3934/math.2023863 |
[5] | Umar Ishtiaq, Fahad Jahangeer, Doha A. Kattan, Manuel De la Sen . Generalized common best proximity point results in fuzzy multiplicative metric spaces. AIMS Mathematics, 2023, 8(11): 25454-25476. doi: 10.3934/math.20231299 |
[6] | Leyla Sağ Dönmez, Abdurrahman Büyükkaya, Mahpeyker Öztürk . Fixed-point results via αji-(DC(PˆE))-contractions in partial ♭-metric spaces. AIMS Mathematics, 2023, 8(10): 23674-23706. doi: 10.3934/math.20231204 |
[7] | Jamshaid Ahmad, Abdullah Shoaib, Irshad Ayoob, Nabil Mlaiki . Common fixed points for (κGm)-contractions with applications. AIMS Mathematics, 2024, 9(6): 15949-15965. doi: 10.3934/math.2024772 |
[8] | Muhammad Waseem Asghar, Mujahid Abbas, Cyril Dennis Enyi, McSylvester Ejighikeme Omaba . Iterative approximation of fixed points of generalized αm-nonexpansive mappings in modular spaces. AIMS Mathematics, 2023, 8(11): 26922-26944. doi: 10.3934/math.20231378 |
[9] | Tahair Rasham, Muhammad Nazam, Hassen Aydi, Abdullah Shoaib, Choonkil Park, Jung Rye Lee . Hybrid pair of multivalued mappings in modular-like metric spaces and applications. AIMS Mathematics, 2022, 7(6): 10582-10595. doi: 10.3934/math.2022590 |
[10] | Amer Hassan Albargi, Jamshaid Ahmad . Fixed point results of fuzzy mappings with applications. AIMS Mathematics, 2023, 8(5): 11572-11588. doi: 10.3934/math.2023586 |
Under addressing global competition, manufacturing companies strive to produce better and cheaper products more quickly. For a complex production system, the design problem is intrinsically a daunting optimization task often involving multiple disciplines, nonlinear mathematical model, and computation-intensive processes during manufacturing process. Here is a reason to develop a high performance algorithm for finding an optimal solution to the engineering design and/or optimization problems. In this paper, a hybrid metaheuristic approach is proposed for solving engineering optimization problems. A genetic algorithm (GA), particle swarm optimization (PSO), and teaching and learning-based optimization (TLBO), called the GA-PSO-TLBO approach, is used and demonstrated for the proposed hybrid metaheuristic approach. Since each approach has its strengths and weaknesses, the GA-PSO-TLBO approach provides an optimal strategy that maintains the strengths as well as mitigates the weaknesses, as needed. The performance of the GA-PSO-TLBO approach is compared with those of conventional approaches such as single metaheuristic approaches (GA, PSO and TLBO) and hybrid metaheuristic approaches (GA-PSO and GA-TLBO) using various types of engineering optimization problems. An additional analysis for reinforcing the performance of the GA-PSO-TLBO approach was also carried out. Experimental results proved that the GA-PSO-TLBO approach outperforms conventional competing approaches and demonstrates high flexibility and efficiency.
In 1922, S. Banach [15] provided the concept of Contraction theorem in the context of metric space. After, Nadler [28] introduced the concept of set-valued mapping in the module of Hausdroff metric space which is one of the potential generalizations of a Contraction theorem. Let (X,d) is a complete metric space and a mapping T:X→CB(X) satisfying
H(T(x),T(y))≤γd(x,y) |
for all x,y∈X, where 0≤γ<1, H is a Hausdorff with respect to metric d and CB(X)={S⊆X:S is closed and bounded subset of X equipped with a metric d}. Then T has a fixed point in X.
In the recent past, Matthews [26] initiate the concept of partial metric spaces which is the classical extension of a metric space. After that, many researchers generalized some related results in the frame of partial metric spaces. Recently, Asadi et al. [4] introduced the notion of an M-metric space which is the one of interesting generalizations of a partial metric space. Later on, Samet et al. [33] introduced the class of mappings which known as (α,ψ)-contractive mapping. The notion of (α,ψ) -contractive mapping has been generalized in metric spaces (see more [10,12,14,17,19,25,29,30,32]).
Throughout this manuscript, we denote the set of all positive integers by N and the set of real numbers by R. Let us recall some basic concept of an M-metric space as follows:
Definition 1.1. [4] Let m:X×X→R+be a mapping on nonempty set X is said to be an M-metric if for any x,y,z in X, the following conditions hold:
(i) m(x,x)=m(y,y)=m(x,y) if and only if x=y;
(ii) mxy≤m(x,y);
(iii) m(x,y)=m(y,x);
(iv) m(x,y)−mxy≤(m(x,z)−mxz)+(m(z,y)−mz,y) for all x,y,z∈X. Then a pair (X,m) is called M-metric space. Where
mxy=min{m(x,x),m(y,y)} |
and
Mxy=max{m(x,x),m(y,y)}. |
Remark 1.2. [4] For any x,y,z in M-metric space X, we have
(i) 0≤Mxy+mxy=m(x,x)+m(y,y);
(ii) 0≤Mxy−mxy=|m(x,x)−m(y,y)|;
(iii) Mxy−mxy≤(Mxz−mxz)+(Mzy−mzy).
Example 1.3. [4] Let (X,m) be an M-metric space. Define mw, ms:X×X→R+ by:
(i)
mw(x,y)=m(x,y)−2mx,y+Mx,y, |
(ii)
ms={m(x,y)−mx,y, if x≠y0, if x=y. |
Then mw and ms are ordinary metrics. Note that, every metric is a partial metric and every partial metric is an M-metric. However, the converse does not hold in general. Clearly every M-metric on X generates a T0 topology τm on X whose base is the family of open M -balls
{Bm(x,ϵ):x∈X, ϵ>0}, |
where
Bm(x,ϵ)={y∈X:m(x,y)<mxy+ϵ} |
for all x∈X, ε>0. (see more [3,4,23]).
Definition 1.4. [4] Let (X,m) be an M-metric space. Then,
(i) A sequence {xn} in (X,m) is said to be converges to a point x in X with respect to τm if and only if
limn→∞(m(xn,x)−mxnx)=0. |
(ii) Furthermore, {xn} is said to be an M-Cauchy sequence in (X,m) if and only if
limn,m→∞(m(xn,xm)−mxnxm), and limn,m→∞(Mxn,xm−mxnxm) |
exist (and are finite).
(iii) An M-metric space (X,m) is said to be complete if every M-Cauchy sequence {xn} in (X,m) converges with respect to τm to a point x∈X such that
limn→∞m(xn,x)−mxnx=0, and limn→∞(Mxn,x−mxnx)=0. |
Lemma 1.5. [4] Let (X,m) be an M-metric space. Then:
(i) {xn} is an M-Cauchy sequence in (X,m) if and only if {xn} is a Cauchy sequence in a metric space (X,mw).
(ii) An M-metric space (X,m) is complete if and only if the metric space (X,mw) is complete. Moreover,
limn→∞mw(xn,x)=0 if and only if (limn→∞(m(xn,x)−mxnx)=0, limn→∞(Mxnx−mxnx)=0). |
Lemma 1.6. [4] Suppose that {xn} convergesto x and {yn} converges to y as n approaches to ∞ in M-metric space (X,m). Then we have
limn→∞(m(xn,yn)−mxnyn)=m(x,y)−mxy. |
Lemma 1.7. [4] Suppose that {xn} converges to xas n approaches to ∞ in M-metric space (X,m).Then we have
limn→∞(m(xn,y)−mxny)=m(x,y)−mxy for all y∈X. |
Lemma 1.8. [4] Suppose that {xn} converges to xand {xn} converges to y as n approaches to ∞ in M-metric space (X,m). Then m(x,y)=mxymoreover if m(x,x)= m(y,y), then x=y.
Definition 1.9. Let α:X×X→[0,∞). A mapping T:X→X is said to be an α-admissible mapping if for all x,y∈X
α(x,y)≥1⇒α(T(x),T(y))≥1. |
Let Ψ be the family of the (c)-comparison functions ψ:R+∪{0}→R+∪{0} which satisfy the following properties:
(i) ψ is nondecreasing,
(ii) ∑∞n=0ψn(t)<∞ for all t>0, where ψn is the n-iterate of ψ (see [7,8,10,11]).
Definition 1.10. [33] Let (X,d) be a metric space and α:X×X→[0,∞). A mapping T:X→X is called (α,ψ)-contractive mapping if for all x,y∈X, we have
α(x,y)d(T(x),T(x))≤ψ(d(x,y)), |
where ψ∈Ψ.
A subset K of an M-metric space X is called bounded if for all x∈K, there exist y∈X and r>0 such that x∈Bm(y,r). Let ¯K denote the closure of K. The set K is closed in X if and only if ¯K=K.
Definition 1.11. [31] Define Hm:CBm(X)×CBm(X)→[0,∞) by
Hm(K,L)=max{∇m(K,L),∇m(L,K)}, |
where
m(x,L)=inf{m(x,y):y∈L} and∇m(L,K)=sup{m(x,L):x∈K}. |
Lemma 1.12. [31] Let F be any nonempty set in M-metric space (X,m), then
x∈¯F if and only if m(x,F)=supa∈F{mxa}. |
Proposition 1.13. [31] Let A,B,C∈CBm(X), then
(i) ∇m(A,A)=supx∈A{supy∈Amxy},
(ii) (∇m(A,B)−supx∈Asupy∈Bmxy)≤(∇m(A,C)−infx∈Ainfz∈Cmxz)+
(∇m(C,B)−infz∈Cinfy∈Bmzy).
Proposition 1.14. [31] Let A,B,C∈CBm(X) followingare hold
(i) Hm(A,A)=∇m(A,A)=supx∈A{supy∈Amxy},
(ii) Hm(A,B)=Hm(B,A),
(iii) Hm(A,B)−supx∈Asupy∈Amxy)≤Hm(A,C)+Hm(B,C)−infx∈Ainfz∈Cmxz−infz∈Cinfy∈Bmzy.
Lemma 1.15. [31] Let A,B∈CBm(X) and h>1.Then for each x∈A, there exist at the least one y∈B such that
m(x,y)≤hHm(A,B). |
Lemma 1.16. [31] Let A,B∈CBm(X) and l>0.Then for each x∈A, there exist at least one y∈B such that
m(x,y)≤Hm(A,B)+l. |
Theorem 1.17. [31] Let (X,m) be a complete M-metric space and T:X→CBm(X). Assume that there exist h∈(0,1) such that
Hm(T(x),T(y))≤hm(x,y), | (1.1) |
for all x,y∈X. Then T has a fixed point.
Proposition 1.18. [31] Let T:X→CBm(X) be a set-valued mapping satisfying (1.1) for all x,y inan M-metric space X. If z∈T(z) for some z in Xsuch that m(x,x)=0 for x∈T(z).
We start with the following definition:
Definition 2.1. Assume that Ψ is a family of non-decreasing functions ϕM:R+→R+ such that
(i) ∑+∞nϕnM(x)<∞ for every x>0 where ϕnM is a nth-iterate of ϕM,
(ii) ϕM(x+y)≤ϕM(x)+ϕM(y) for all x,y∈R+,
(iii) ϕM(x)<x, for each x>0.
Remark 2.2. If ∑αn|n=∞ =0 is a convergent series with positive terms then there exists a monotonic sequence (βn)|n=∞ such that βn|n=∞=∞ and ∑αnβn|n=∞=0 converges.
Definition 2.3. Let (X,m) be an M-metric pace. A self mapping T:X→X is called (α∗,ϕM)-contraction if there exist two functions α∗:X×X→[0,∞) and ϕM∈Ψ such that
α∗(x,y)m(T(x),T(y))≤ϕM(m(x,y)), |
for all x,y∈X.
Definition 2.4. Let (X,m) be an M-metric space. A set-valued mapping T:X→CBm(X) is said to be (α∗,ϕM)-contraction if for all x,y∈X, we have
α∗(x,y)Hm(T(x),T(x))≤ϕM(m(x,y)), | (2.1) |
where ϕM∈Ψ and α∗:X×X→[0,∞).
A mapping T is called α∗-admissible if
α∗(x,y)≥1⇒α∗(a1,b1)≥1 |
for each a1∈T(x) and b1∈T(y).
Theorem 2.5. Let (X,m) be a complete M-metric space.Suppose that (α∗,ϕM) contraction and α∗-admissible mapping T:X→CBm(X)satisfies the following conditions:
(i) there exist x0∈X such that α∗(x0,a1)≥1 for each a1∈T(x0),
(ii) if {xn}∈X is a sequence such that α∗(xn,xn+1)≥1 for all n and {xn}→x∈X as n→∞, then α∗(xn,x)≥1 for all n∈N. Then T has a fixed point.
Proof. Let x1∈T(x0) then by the hypothesis (i) α∗(x0,x1)≥1. From Lemma 1.16, there exist x2∈T(x1) such that
m(x1,x2)≤Hm(T(x0),T(x1))+ϕM(m(x0,x1)). |
Similarly, there exist x3∈T(x2) such that
m(x2,x3)≤Hm(T(x1),T(x2))+ϕ2M(m(x0,x1)). |
Following the similar arguments, we obtain a sequence {xn}∈X such that there exist xn+1∈T(xn) satisfying the following inequality
m(xn,xn+1)≤Hm(T(xn−1),T(xn))+ϕnM(m(x0,x1)). |
Since T is α∗-admissible, therefore α∗(x0,x1)≥1⇒α∗(x1,x2)≥1. Using mathematical induction, we get
α∗(xn,xn+1)≥1. | (2.2) |
By (2.1) and (2.2), we have
m(xn,xn+1)≤Hm(T(xn−1),T(xn))+ϕnM(m(x0,x1))≤α∗(xn,xn+1)Hm(T(xn−1),T(xn))+ϕnM(m(x0,x1))≤ϕM(m(xn−1,xn))+ϕnM(m(x0,x1))=ϕM[(m(xn−1,xn))+ϕn−1M(m(x0,x1))]≤ϕM[Hm(T(xn−2),T(xn−1))+ϕn−1M(m(x0,x1))]≤ϕM[α∗(xn−1,xn)Hm(T(xn−1),T(xn))+ϕn−1M(m(x0,x1))]≤ϕM[ϕM(m(xn−2,xn−1))+ϕn−1M(m(x0,x1))+ϕn−1M(m(x0,x1))]≤ϕ2M(m(xn−2,xn−1))+2ϕnM(m(x0,x1)).... |
m(xn,xn+1)≤ϕnM(m(x0,x1))+nϕnM(m(x0,x1))m(xn,xn+1)≤(n+1)ϕnM(m(x0,x1)). |
Let us assume that ϵ>0, then there exist n0∈N such that
∑n≥n0(n+1)ϕnM(m(x0,x1))<ϵ. |
By the Remarks (1.2) and (2.2), we get
limn→∞m(xn,xn+1)=0. |
Using the above inequality and (m2), we deduce that
limn→∞m(xn,xn)=limn→∞min{m(xn,xn),m(xn+1,xn+1)}=limn→∞mxnxn+1≤limn→∞m(xn,xn+1)=0. |
Owing to limit, we have limn→∞m(xn,xn)=0,
limn,m→∞mxnxm=0. |
Now, we prove that {xn} is M-Cauchy in X. For m,n in N with m>n and using the triangle inequality of an M-metric we get
m(xn,xm)−mxnxm≤m(xn,xn+1)−mxnxn+1+m(xn+1,xm)−mxn+1xm≤m(xn,xn+1)−mxnxn+1+m(xn+1,xn+2)−mxn+1xn+1+m(xn+2,xm)−mxn+2xm≤m(xn,xn+1)−mxnxn+1+m(xn+1,xn+2)−mxn+1xn+2+⋅⋅⋅+m(xm−1,xm)−mxm−1xm≤m(xn,xn+1)+m(xn+1,xn+2)+⋅⋅⋅+m(xm−1,xm)=m−1∑r=nm(xr,xr+1)≤m−1∑r=n(r+1)ϕrM(m(x0,x1))≤m−1∑r≥n0(r+1)ϕrM(m(x0,x1))≤m−1∑r≥n0(r+1)ϕrM(m(x0,x1))<ϵ. |
m(xn,xm)−mxnxm→0, as n→∞, we obtain limm,n→∞(Mxnxm−mxnxm)=0. Thus {xn} is a M-Cauchy sequence in X. Since (X,m) is M-complete, there exist x⋆∈X such that
limn→∞(m(xn,x⋆)−mxnx⋆)=0 andlimn→∞(Mxnx⋆−mxnx⋆)=0. |
Also, limn→∞m(xn,xn)=0 gives that
limn→∞m(xn,x⋆)=0 and limn→∞Mxnx⋆=0, | (2.3) |
limn→∞{max(m(xn,x⋆),m(x⋆,x⋆))}=0, |
which implies that m(x⋆,x⋆)=0 and hence we obtain mx⋆T(x⋆)=0. By using (2.1) and (2.3) with
limn→∞α∗(xn,x⋆)≥1. |
Thus,
limn→∞Hm(T(xn),T(x⋆))≤limn→∞ϕM(m(xn,x⋆))≤limn→∞m(xn,x⋆). |
limn→∞Hm(T(xn),T(x⋆))=0. | (2.4) |
Now from (2.3), (2.4), and xn+1∈T(xn), we have
m(xn+1,T(x⋆))≤Hm(T(xn),T(x⋆))=0. |
Taking limit as n→∞ and using (2.4), we obtain that
limn→∞m(xn+1,T(x⋆))=0. | (2.5) |
Since mxn+1T(x⋆)≤m(xn+1,T(x⋆)) which gives
limn→∞mxn+1T(x⋆)=0. | (2.6) |
Using the condition (m4), we obtain
m(x⋆,T(x⋆))−supy∈T(x⋆)mx⋆y≤m(x⋆,T(x⋆))−mx⋆,T(x⋆)≤m(x⋆,xn+1)−mx⋆xn+1+m(xn+1,T((x⋆))−mxn+1T(x⋆). |
Applying limit as n→∞ and using (2.3) and (2.6), we have
m(x⋆,T(x⋆))≤supy∈T(x⋆)mx⋆y. | (2.7) |
From (m2), mx⋆y≤m(x⋆y) for each y∈T(x⋆) which implies that
mx⋆y−m(x⋆,y)≤0. |
Hence,
sup{mx⋆y−m(x⋆,y):y∈T(x⋆)}≤0. |
Then
supy∈T(x⋆)mx⋆y−infy∈T(x⋆)m(x⋆,y)≤0. |
Thus
supy∈T(x⋆)mx⋆y≤m(x⋆,T(x⋆)). | (2.8) |
Now, from (2.7) and (2.8), we obtain
m(T(x⋆),x⋆)=supy∈T(x⋆)mx⋆y. |
Consequently, owing to Lemma (1.12), we have x⋆∈¯T(x⋆)=T(x⋆).
Corollary 2.6. Let (X,m) be a complete M-metric space and anself mapping T:X→X an α∗-admissible and (α∗,ϕM)-contraction mapping. Assume that thefollowing properties hold:
(i) there exists x0∈X such that α∗(x0,T(x0))≥1,
(ii) either T is continuous or for any sequence {xn}∈X with α∗(xn,xn+1)≥1 for all n∈N and {xn}→x as n → ∞, we have α∗(xn,x)≥1 for all n∈N. Then T has a fixed point.
Some fixed point results in ordered M-metric space.
Definition 2.7. Let (X,⪯) be a partially ordered set. A sequence {xn}⊂X is said to be non-decreasing if xn⪯xn+1 for all n.
Definition 2.8. [16] Let F and G be two nonempty subsets of partially ordered set (X,⪯). The relation between F and G is defined as follows: F≺1G if for every x∈F, there exists y∈G such that x⪯y.
Definition 2.9. Let (X,m,⪯) be a partially ordered set on M-metric. A set-valued mapping T:X→CBm(X) is said to be ordered (α∗,ϕM)-contraction if for all x,y∈X, with x⪯y we have
Hm(T(x),T(y))≤ϕM(m(x,y)) |
where ϕM∈Ψ. Suppose that α∗:X×X→[0,∞) is defined by
α∗(x,y)={1 if Tx≺1Ty0 otherwise. |
A mapping T is called α∗-admissible if
α(x,y)≥1⇒α∗(a1,b1)≥1, |
for each a1∈T(x) and b1∈T(y).
Theorem 2.10. Let (X,m,⪯) be a partially orderedcomplete M-metric space and T:X→CBm(X) an α∗-admissible ordered (α∗,ϕM)-contraction mapping satisfying the following conditions:
(i) there exist x0∈X such that {x0}≺1{T(x0)}, α∗(x0,a1)≥1 for each a1∈T(x0),
(ii) for every x,y∈X, x⪯y implies T(x)≺1T(y),
(iii) If {xn}∈X is a non-decreasing sequence such that xn⪯xn+1 for all n and {xn}→x∈X as n →∞ gives xn⪯x for all n∈N. Then T has a fixed point.
Proof. By assumption (i) there exist x1∈T(x0) such that x0⪯x1 and α∗(x0,x1)≥1. By hypothesis (ii), T(x0)≺1T(x1). Let us assume that there exist x2∈T(x1) such that x1⪯x2 and we have the following
m(x1,x2)≤Hm(T(x0),T(x1))+ϕM(m(x0,x1)). |
In the same way, there exist x3∈T(x2) such that x2⪯x3 and
m(x2,x3)≤Hm(T(x1),T(x2))+ϕ2M(m(x0,x1)). |
Following the similar arguments, we have a sequence {xn}∈X and xn+1∈T(xn) for all n≥0 satisfying x0⪯x1⪯x2⪯x3⪯...xn⪯xn+1. The proof is complete follows the arguments given in Theorem 2.5.
Example 2.11. Let X=[16,1] be endowed with an M -metric given by m(x,y)=x+y2. Define T:X→CBm(X) by
T(x)={{12x+16,14}, if x=16{x2,x3}, if 14≤x≤13{23,56}, if 12≤x≤1. |
Define a mapping α∗:X×X→[0,∞) by
α∗(x,y)={1 if x,y∈[14,13]0 otherwise. |
Let ϕM:R+→R+ be given by ϕM(t)=1710 where ϕM∈Ψ, for x,y∈X. If x=16, y=14 then m(x,y)=524, and
Hm(T(x),T(y))=Hm({312,14},{18,112})=max(∇m({312,14},{18,112}),∇m({18,112},{312,14}))=max{316,212}=316≤ϕM(t)m(x,y). |
If x=13, y=12 then m(x,y)=512, and
Hm(T(x),T(y))=Hm({16,19},{23,1})=max(∇m({16,19},{23,1}),∇m({23,1},{16,19}))=max{1736,718}=1736≤ϕM(t)m(x,y). |
If x=16, y=1, then m(x,y)=712 and
Hm(T(x),T(y))=Hm({312,14},{23,56})=max(∇m({312,14},{23,56}),∇m({23,56},{312,14}))=max{1124,1324}=1324≤ϕM(t)m(x,y). |
In all cases, T is (α∗,ϕM)-contraction mapping. If x0=13, then T(x0)={x2,x3}.Therefore α∗(x0,a1)≥1 for every a1∈T(x0). Let x,y∈X be such that α∗(x,y)≥1, then x,y∈[x2,x3] and T(x)={x2,x3} and T(y)= {x2,x3} which implies that α∗(a1,b1)≥1 for every a1∈T(x) and b1∈T(x). Hence T is α∗-admissble.
Let {xn}∈X be a sequence such that α∗(xn,xn+1)≥1 for all n in N and xn converges to x as n converges to ∞, then xn∈[x2,x3]. By definition of α∗ -admissblity, therefore x∈[x2,x3] and hence α∗(xn,x)≥1. Thus all the conditions of Theorem 2.3 are satisfied. Moreover, T has a fixed point.
Example 2.12. Let X={(0,0),(0,−15),(−18,0)} be the subset of R2 with order ⪯ defined as: For (x1,y1),(x2,y2)∈X, (x1,y1)⪯(x2,y2) if and only if x1≤x2, y1≤y2. Let m:X×X→R+ be defined by
m((x1,y1),(x2,y2))=|x1+x22|+|y1+y22|, for x=(x1,y1), y=(x2,y2)∈X. |
Then (X,m) is a complete M-metric space. Let T:X→CBm(X) be defined by
T(x)={{(0,0)}, if x=(0,0),{(0,0),(−18,0)}, if x∈(0,−15){(0,0)}, if x∈(−18,0). |
Define a mapping α∗:X×X→[0,∞) by
α∗(x,y)={1 if x,y∈X0 otherwise. |
Let ϕM:R+→R+ be given by ϕM(t)=12. Obviously, ϕM∈Ψ. For x,y∈X,
if x=(0,−15) and y=(0,0), then Hm(T(x),T(y))=0 and m(x,y)=110 gives that
Hm(T(x),T(y))=Hm({(0,0),(−18,0)},{(0,0)})=max(∇m({(0,0),(−18,0)},{(0,0)}),∇m({(0,0)},{(0,0),(−18,0)}))=max{0,0}=0≤ϕM(t)m(x,y). |
If x=(−18,0) and y=(0,0) then Hm(T(x),T(y))=0, and m(x,y)=116 implies that
Hm(T(x),T(y))≤ϕM(t)m(x,y). |
If x=(0,0) and y=(0,0) then Hm(T(x),T(y))=0, and m(x,y)=0 gives
Hm(T(x),T(y))≤ϕM(t)m(x,y). |
If x=(0,−15) and y=(0,−15) then Hm(T(x),T(y))=0, and m(x,y)=15 implies that
Hm(T(x),T(y))≤ϕM(t)m(x,y). |
If x=(0,−18) and y=(0,−18) then Hm(T(x),T(y))=0, and m(x,y)=18 gives that
Hm(T(x),T(y))≤ϕM(t)m(x,y). |
Thus all the condition of Theorem 2.10 satisfied. Moreover, (0,0) is the fixed point of T.
In this section, we present an application of our result in homotopy theory. We use the fixed point theorem proved for set-valued (α∗,ϕM)-contraction mapping in the previous section, to establish the result in homotopy theory. For further study in this direction, we refer to [6,35].
Theorem 3.1. Suppose that (X,m) is a complete M-metricspace and A and B are closed and open subsets of X respectively, suchthat A⊂B. For a,b∈R, let T:B×[a,b]→CBm(X) be aset-valued mapping satisfying the following conditions:
(i) x∉T(y,t) for each y∈B/Aand t∈[a,b],
(ii) there exist ϕM∈Ψ and α∗:X×X→[0,∞) such that
α∗(x,y)Hm(T(x,t),T(y,t))≤ϕM(m(x,y)), |
for each pair (x,y)∈B×B and t∈[a,b],
(iii) there exist a continuous function Ω:[a,b]→R such that for each s,t∈[a,b] and x∈B, we get
Hm(T(x,s),T(y,t))≤ϕM|Ω(s)−Ω(t)|, |
(iv) if x⋆∈T(x⋆,t),then T(x⋆,t)={x⋆},
(v) there exist x0 in X such that x0∈T(x0,t),
(vi) a function ℜ:[0,∞)→[0,∞) defined by ℜ(x)=x−ϕM(x) is strictly increasing and continuous if T(.,t⊺ has a fixed point in B for some t^{\intercal }\in \left[a, b\right], then T\left(., t\right) has afixed point in A for all t\in \left[a, b\right]. Moreover, for a fixed t\in \left[a, b\right] , fixed point is unique provided that \phi_{M}\left(t\right) = \frac{1}{2}t where t > 0.
Proof. Define a mapping \alpha _{\ast }:X\times X\rightarrow \left[0, \infty \right) by
\alpha _{\ast }\left( x, y\right) = \left \{ \begin{array}{l} 1 \ \ \ \ \text{ if }x\in T\left( x, t\right) , \ y\in T\left( y, t\right) \\ \ \\ 0 \ \ \ \ \ \ \ \text{otherwise.} \end{array} \right. |
We show that T is \alpha _{\ast } -admissible. Note that \alpha _{\ast }\left(x, y\right) \geq 1 implies that x\in T\left(x, t\right) and y\in T\left(y, t\right) for all t\in \left[a, b\right] . By hypothesis \left(iv\right), T\left(x, t\right) = \left \{ x\right \} and T\left(y, t\right) = \left \{ y\right \}. It follows that T is \alpha _{\ast } -admissible. By hypothesis \left(v\right), there exist x_{0}\in X such that x_{0}\in \left(x_{0}, t\right) for all t , that is \alpha _{\ast }\left(x_{0}, x_{0}\right) \geq 1 . Suppose that \alpha _{\ast }\left(x_{n}, x_{n+1}\right) \geq 1 for all n and x_{n} converges to q as n approaches to \infty and x_{n}\in T\left(x_{n}, t\right) and x_{n+1}\in T\left(x_{n+1}, t\right) for all n and t\in \left[a, b\right] which implies that q\in T\left(q, t\right) and thus \alpha _{\ast }\left(x_{n}, q\right) \geq 1. Set
D = \left \{ t\in \left[ a, b\right] : \ x\in T\left( x, t\right) \text{ for }x\in A\right \} . |
So T\left(., t^{\intercal }\right) has a fixed point in B for some t^{\intercal }\in \left[a, b\right] , there exist x\in B such that x\in T\left(x, t\right). By hypothesis \left(i\right) x\in T\left(x, t\right) for t\in \left[a, b\right] and x\in A so D\neq \phi . Now we now prove that D is open and close in \left[a, b\right]. Let t_{0}\in D and x_{0}\in A with x_{0}\in T\left(x_{0}, t_{0}\right). Since A is open subset of X , \overline{B_{m}\left(x_{0}, r\right) } \subseteq A for some r > 0 . For \epsilon = r+m_{xx_{0}}-\phi \left(r+m_{xx_{0}}\right) and a continuous function \Omega on \left[a, b \right] , there exist \delta > 0 such that
\phi _{M}\left \vert \Omega \left( t\right) -\Omega \left( t_{0}\right) \right \vert < \epsilon \text{ for all }t\in \left( t_{0}-\delta , t_{0}+\delta \right) . |
If t\in \left(t_{0}-\delta, t_{0}+\delta \right) for x\in B_{m}\left(x_{0}, r\right) = \left \{ x\in X:m\left(x_{0}\, , x\right) \leq m_{x_{0}x}+r\right \} and l\in T\left(x, t\right), we obtain
\begin{eqnarray*} m\left( l, x_{0}\right) & = &m\left( T\left( x, t\right) , x_{0}\right) \\ & = &H_{m}\left( T\left( x, t\right) , T\left( x_{0}, t_{0}\right) \right) . \end{eqnarray*} |
Using the condition \left(iii\right) of Proposition 1.13 and Proposition 1.18, we have
\begin{equation} m\left( l, x_{0}\right) \leq H_{m}\left( T\left( x, t\right) , T\left( x_{0}, t_{0}\right) \right) +H_{m}\left( T\left( x, t\right) , T\left( x_{0}, t_{0}\right) \right) \end{equation} | (2.9) |
as x\in T\left(x_{0}, t_{0}\right) and x\in B_{m}\left(x_{0}, r\right) \subseteq A\subseteq B , t_{0}\in \left[a, b\right] with \alpha _{\ast }\left(x_{0}, x_{0}\right) \geq 1. By hypothesis \left(ii\right) , \left(iii\right) and \left(2.9\right)
\begin{eqnarray*} m\left( l, x_{0}\right) &\leq &\phi _{M}\left \vert \Omega \left( t\right) -\Omega \left( t_{0}\right) \right \vert +\alpha _{\ast }\left( x_{0}, x_{0}\right) H_{m}\left( T\left( x, t\right) , T\left( x_{0}, t_{0}\right) \right) \\ &\leq &\phi _{M}\left \vert \Omega \left( t\right) -\Omega \left( t_{0}\right) \right \vert +\phi _{M}\left( m\left( x, x_{0}\right) \right) \\ &\leq &\phi _{M}\left( \epsilon \right) +\phi _{M}\left( m_{xx_{0}}+r\right) \\ &\leq &\phi _{M}\left( r+m_{xx_{0}}-\phi _{M}\left( r+m_{xx_{0}}\right) \right) +\phi _{M}\left( m_{xx_{0}}+r\right) \\ & < &r+m_{xx_{0}}-\phi _{M}\left( r+m_{xx_{0}}\right) +\phi _{M}\left( m_{xx_{0}}+r\right) = r+m_{xx_{0}}. \end{eqnarray*} |
Hence l\in \overline{B_{m}\left(x_{0}, r\right) } and thus for each fixed t\in \left(t_{0}-\delta, t_{0}+\delta \right), we obtain T\left(x, t\right) \subset \overline{B_{m}\left(x_{0}, r\right) } therefore T: \overline{B_{m}\left(x_{0}, r\right) }\rightarrow CB_{m}\left(\overline{ B_{m}\left(x_{0}, r\right) }\right) satisfies all the assumption of Theorem \left(3.1\right) and T\left(., t\right) has a fixed point \overline{B_{m}\left(x_{0}, r\right) } = B_{m}\left(x_{0}, r\right) \subset B . But by assumption of \left(i\right) this fixed point belongs to A . So \left(t_{0}-\delta, t_{0}+\delta \right) \subseteq D, thus D is open in \left[a, b\right]. Next we prove that D is closed. Let a sequence \left \{ t_{n}\right \} \in D with t_{n} converges to t_{0}\in \left[a, b \right] as n approaches to \infty. We will prove that t_{0} is in D .
Using the definition of D, there exist \left \{ t_{n}\right \} in A such that x_{n}\in T\left(x_{n}, t_{n}\right) for all n . Using Assumption \left(iii\right) – \left(v\right), and the condition \left(iii\right) of Proposition 1.13, and an outcome of the Proposition 1.18, we have
\begin{eqnarray*} m\left( x_{n}, x_{m}\right) &\leq &H_{m}\left( T\left( x_{n}, t_{n}\right) , T\left( x_{m}, t_{m}\right) \right) \\ &\leq &H_{m}\left( T\left( x_{n}, t_{n}\right) , T\left( x_{n}, t_{m}\right) \right) +H_{m}\left( T\left( x_{n}, t_{m}\right) , T\left( x_{m}, t_{m}\right) \right) \\ &\leq &\phi _{M}\left \vert \Omega \left( t_{n}\right) -\Omega \left( t_{m}\right) \right \vert +\alpha _{\ast }\left( x_{n}, x_{m}\right) H_{m}\left( T\left( x_{n}, t_{m}\right) , T\left( x_{m}, t_{m}\right) \right) \\ &\leq &\phi _{M}\left \vert \Omega \left( t_{n}\right) -\Omega \left( t_{m}\right) \right \vert +\phi _{M}\left( m\left( x_{n}, x_{m}\right) \right) \\ &\Rightarrow & \\ m\left( x_{n}, x_{m}\right) -\phi _{M}\left( m\left( x_{n}, x_{m}\right) \right) &\leq &\phi _{M}\left \vert \Omega \left( t_{n}\right) -\Omega \left( t_{m}\right) \right \vert \\ &\Rightarrow & \\ \Re \left( m\left( x_{n}, x_{m}\right) \right) &\leq &\phi _{M}\left \vert \Omega \left( t_{n}\right) -\Omega \left( t_{m}\right) \right \vert \\ \Re \left( m\left( x_{n}, x_{m}\right) \right) \, & < &\left \vert \Omega \left( t_{n}\right) -\Omega \left( t_{m}\right) \right \vert \\ m\left( x_{n}, x_{m}\right) & < &\frac{1}{\Re }\left \vert \Omega \left( t_{n}\right) -\Omega \left( t_{m}\right) \right \vert . \end{eqnarray*} |
So, continuity of \frac{1}{\Re }, \Re and convergence of \left \{ t_{n}\right \}, taking the limit as m, n\rightarrow \infty in the last inequality, we obtain that
\lim\limits_{m, n\rightarrow \infty }m\left( x_{n}, x_{m}\right) = 0. |
Sine m_{x_{n}x_{m}}\leq m\left(x_{n}, x_{m}\right), therefore
\lim\limits_{m, n\rightarrow \infty }m_{x_{n}x_{m}} = 0. |
Thus, we have \lim_{n\rightarrow \infty }m\left(x_{n}, x_{n}\right) = 0 = \lim_{m\rightarrow \infty }m\left(x_{m}, x_{m}\right) . Also,
\lim\limits_{m, n\rightarrow \infty }\left( m\left( x_{n}, x_{m}\right) -m_{x_{n}x_{m}}\right) = 0, \ \lim\limits_{m, n\rightarrow \infty }\left( M_{x_{n}x_{m}}-m_{x_{n}x_{m}}\right) . |
Hence \left \{ x_{n}\right \} is an M -Cauchy sequence. Using Definition 1.4, there exist x^{\ast } in X such that
\lim\limits_{n\rightarrow \infty }\left( m\left( x_{n}, x^{\ast }\right) -m_{x_{n}x^{\ast }}\right) = 0\text{ and }\lim\limits_{n\rightarrow \infty }\left( M_{x_{n}x^{\ast }}-m_{x_{n}x^{\ast }}\right) = 0. |
As \lim_{n\rightarrow \infty }m\left(x_{n}, x_{n}\right) = 0 , therefore
\lim\limits_{n\rightarrow \infty }m\left( x_{n}, x^{\ast }\right) = 0\text{ and } \lim\limits_{n\rightarrow \infty }M_{x_{n}x^{\ast }} = 0. |
Thus, we have m\left(x, x^{\ast }\right) = 0. We now show that x^{\ast }\in T\left(x^{\ast }, t^{^{\ast }}\right). Note that
\begin{eqnarray*} m\left( x_{n}, T\left( x^{\ast }, t^{^{\ast }}\right) \right) &\leq &H_{m}\left( T\left( x_{n}, t_{n}\right) , T\left( x^{\ast }, t^{^{\ast }}\right) \right) \\ &\leq &H_{m}\left( T\left( x_{n}, t_{n}\right) , T\left( x_{n}, t^{^{\ast }}\right) \right) +H_{m}\left( T\left( x_{n}, t^{^{\ast }}\right) , T\left( x^{\ast }, t^{^{\ast }}\right) \right) \\ &\leq &\phi _{M}\left \vert \Omega \left( t_{n}\right) -\Omega \left( t^{^{\ast }}\right) \right \vert +\alpha _{\ast }\left( x_{n}, t^{^{\ast }}\right) H_{m}\left( T\left( x_{n}, t^{^{\ast }}\right) , T\left( x^{\ast }, t^{^{\ast }}\right) \right) \\ &\leq &\phi _{M}\left \vert \Omega \left( t_{n}\right) -\Omega \left( t^{^{\ast }}\right) \right \vert +\phi _{M}\left( m\left( x_{n}, t^{^{\ast }}\right) \right) . \end{eqnarray*} |
Applying the limit n\rightarrow \infty in the above inequality, we have
\lim\limits_{n\rightarrow \infty }m\left( x_{n}, T\left( x^{\ast }, t^{^{\ast }}\right) \right) = 0. |
Hence
\begin{equation} \lim\limits_{n\rightarrow \infty }m\left( x_{n}, T\left( x^{\ast }, t^{^{\ast }}\right) \right) = 0. \end{equation} | (2.10) |
Since m\left(x^{\ast }, x^{\ast }\right) = 0, we obtain
\begin{equation} \sup\limits_{y\in T\left( x^{\ast }, t^{^{\ast }}\right) }m_{x^{\ast }y} = \sup\limits_{y\in T\left( x^{\ast }, t^{^{\ast }}\right) }\min \left \{ m\left( x^{\ast }, x^{\ast }\right) , m\left( y, y\right) \right \} = 0. \end{equation} | (2.11) |
From above two inequalities, we get
m\left( x^{\ast }, T\left( x^{\ast }, t^{^{\ast }}\right) \right) = \sup\limits_{y\in T\left( x^{\ast }, t^{^{\ast }}\right) }m_{x^{\ast }y}. |
Thus using Lemma 1.12 we get x^{\ast }\in T\left(x^{\ast }, t^{^{\ast }}\right). Hence x^{\ast }\in A. Thus x^{\ast }\in D and D is closed in \left[a, b\right], D = \left[a, b\right] and D is open and close in \left[a, b\right]. Thus T\left(., t\right) has a fixed point in A for all t\in \left[a, b\right]. For uniqueness, t\in \left[a, b\right] is arbitrary fixed point, then there exist x\in A such that x\in T\left(x, t\right) . Assume that y is an other point of T\left(x, t\right) , then by applying condition 4, we obtain
\begin{eqnarray*} m\left( x, y\right) & = &H_{m}\left( T\left( x, t\right) , T\left( y, t\right) \right) \\ &\leq &\alpha _{M}\left( x, y\right) H_{m}\left( T\left( x, t\right) , T\left( y, t\right) \right) \leq \phi _{M}\left( m\left( x, y\right) \right) . \end{eqnarray*} |
For \phi _{M}\left(t\right) = \frac{1}{2}t, where t > 0, the uniqueness follows.
In this section we will apply the previous theoretical results to show the existence of solution for some integral equation. For related results (see [13,20]). We see for non-negative solution of \left(3.1\right) in X = C\left(\left[0, \delta \right], \mathbb{R} \right). Let X = C\left(\left[0, \delta \right], \mathbb{R} \right) be a set of continuous real valued functions defined on \left[0, \delta \right] which is endowed with a complete M -metric given by
m\left( x, y\right) = \sup\limits_{t\in \left[ 0, \delta \right] }\left( \left \vert \frac{x\left( t\right) +x\left( t\right) }{2}\right \vert \right) \text{ for all }x, y\in X. |
Consider an integral equation
\begin{equation} v_{1}\left( t\right) = \rho \left( t\right) +\int_{0}^{\delta }h\left( t, s\right) J\left( s, v_{1}\left( s\right) \right) ds\text{ for all }0\leq t\leq \delta . \end{equation} | (3.1) |
Define g:X\rightarrow X by
g\left( x\right) \left( t\right) = \rho \left( t\right) +\int_{0}^{\delta }h\left( t, s\right) J\left( s, x\left( s\right) \right) ds |
where
(i) for \delta > 0 , \ \left(a\right) J:\left[0, \delta \right] \times \mathbb{R} \rightarrow \mathbb{R}, \left(b\right) h:\left[0, \delta \right] \times \left[0, \delta \right] \rightarrow \left[0, \infty \right), \left(c\right) \rho : \left[0, \delta \right] \rightarrow \mathbb{R} are all continuous functions
(ii) Assume that \sigma :X\times X\rightarrow \mathbb{R} is a function with the following properties,
(iii) \sigma \left(x, y\right) \geq 0 implies that \sigma \left(T\left(x\right), T\left(y\right) \right) \geq 0,
(iv) there exist x_{0}\in X such that \sigma \left(x_{0}, T\left(x_{0}\right) \right) \geq 0,
(v) if \left \{ x_{n}\right \} \in X is a sequence such that \sigma \left(x_{n}, x_{n+1}\right) \geq 0 for all n\in \mathbb{N} and x_{n}\rightarrow x as n\rightarrow \infty, then \sigma \left(x, T\left(x\right) \right) \geq 0
(vi)
\sup\limits_{t\in \left[ 0, \delta \right] }\int_{0}^{\delta }h\left( t, s\right) ds\leq 1 |
where t\in \left[0, \delta \right] , s\in \mathbb{R},
\left(vii\right) there exist \phi _{M}\in \Psi , \sigma \left(y, T\left(y\right) \right) \geq 1 and \sigma \left(x, T\left(x\right) \right) \geq 1 such that for each t\in \left[0, \delta \right], we have
\begin{equation} |J\left( s, x\left( t\right) \right) +J\left( s, y\left( t\right) \right) |\leq \phi _{M}\left( \left \vert x+y\right \vert \right) . \end{equation} | (3.3) |
Theorem 4.1. Under the assumptions \left(i\right) -\left(vii\right) theintegral Eq \left(3.1\right) has a solution in \left \{ X = C\left(\left[0, \delta \right], \mathbb{R} \right) \mathit{\text{ for all }}t\in \left[0, \delta \right] \right \}.
Proof. Using the condition \left(vii\right) , we obtain that
\begin{eqnarray*} m\left( g\left( x\right) , g\left( y\right) \right) & = &\left \vert \frac{ g\left( x\right) \left( t\right) +g\left( y\right) \left( t\right) }{2} \right \vert = \left \vert \int_{0}^{\delta }h\left( t, s\right) \left[ \frac{ J\left( s, x\left( s\right) \right) +J\left( s, y\left( s\right) \right) }{2} \right] ds\right \vert \\ &\leq &\int_{0}^{\delta }h\left( t, s\right) \left \vert \frac{J\left( s, x\left( s\right) \right) +J\left( s, y\left( s\right) \right) }{2} \right \vert ds \\ &\leq &\int_{0}^{\delta }h\left( t, s\right) \left[ \phi _{M}\left \vert \frac{ x\left( s\right) +y\left( s\right) }{2}\right \vert \right] ds \\ &\leq &\left( \sup\limits_{t\in \left[ 0, \delta \right] }\int_{0}^{\delta }h\left( t, s\right) ds\right) \left( \phi _{M}\left \vert \frac{x\left( s\right) +y\left( s\right) }{2}\right \vert \right) \\ &\leq &\phi _{M}\left( \left \vert \frac{x\left( s\right) +y\left( s\right) }{ 2}\right \vert \right) \end{eqnarray*} |
m\left( g\left( x\right) , g\left( y\right) \right) \leq \phi \left( m\left( x, y\right) \right) |
Define \alpha _{\ast }:X\times X\rightarrow \left[0, +\infty \right) by
\alpha _{\ast }\left( x, y\right) = \left \{ \begin{array}{l} 1 \ \ \ \ \text{ if }\sigma \left( x, y\right) \geq 0 \\ \ \\ 0 \ \ \ \ \ \ \ \text{otherwise} \end{array} \right. |
which implies that
m\left( g\left( x\right) , g\left( y\right) \right) \leq \phi _{M}\left( m\left( x, y\right) \right) . |
Hence all the assumption of the Corollary 2.6 are satisfied, the mapping g has a fixed point in X = C\left(\left[0, \delta \right], \mathbb{R} \right) which is the solution of integral Eq \left(3.1\right).
In this study we develop some set-valued fixed point results based on \left(\alpha _{\ast }, \phi _{M}\right) -contraction mappings in the context of M -metric space and ordered M -metric space. Also, we give examples and applications to the existence of solution of functional equations and homotopy theory.
The authors declare that they have no competing interests.
[1] |
M. Gen, Y. Yun, Soft computing approach for reliability optimization: State-of-the-art survey, Rel. Eng. and Sys. Saf., 91 (2006), 1008-1026. https://doi.org/10.1016/j.ress.2005.11.053 doi: 10.1016/j.ress.2005.11.053
![]() |
[2] |
G. Wang, S. Shan, Review of metamodeling techniques in support of engineering design optimization, Trans. ASME, 129 (2007), 370-380. https://doi.org/10.1115/1.2429697 doi: 10.1115/1.2429697
![]() |
[3] |
H. M. Amir, T. Hasegawa, Nonlinear mixed-discrete structural optimization, J. Struct. Eng., 115 (1989), 626-646. https://doi.org/10.1061/(ASCE)0733-9445(1989)115:3(626) doi: 10.1061/(ASCE)0733-9445(1989)115:3(626)
![]() |
[4] |
E. Sandgren, Nonlinear integer and discrete programming in mechanical design optimization, ASME J. Mech. Des., 112 (1990), 223-229. https://doi.org/10.1115/1.2912596 doi: 10.1115/1.2912596
![]() |
[5] |
J. F. Fu, R. G. Fenton, W. L. Cleghorn, A mixed integer-discrete-continuous programming method and its applications to engineering design optimization, Eng. Opt., 17 (1991), 263-280. https://doi.org/10.1080/03052159108941075 doi: 10.1080/03052159108941075
![]() |
[6] | W. Kuo, V. R. Prasad, F. Tillman, C. L. Hwang, Optimal reliability design: Fundamentals and applications, Cambridge University Press, 2001. |
[7] | N. G. Yarushkina, Genetic algorithms for engineering optimization: Theory and practice, in Proceedings of the 2002 IEEE International Conference on Artificial Intelligence System (ICAIS'02), (2002), 357-362. https://doi:10.1109/ICAIS.2002.1048127 |
[8] |
C. -Y. Lin, P. Hajela, Genetic algorithms in optimization problems with discrete and integer design variables, Eng. Opt., 19 (1992), 309-327. http://doi.org/10.1080/03052159208941234 doi: 10.1080/03052159208941234
![]() |
[9] |
S. J. Wu, P. T. Chow, Genetic algorithms for nonlinear mixed discrete-integer optimization problems via meta-genetic parameter optimization, Eng. Opt., 24 (1995), 137-159. http://doi.org/10.1080/03052159508941187 doi: 10.1080/03052159508941187
![]() |
[10] |
T. Yokota, T. Taguchi, M. Gen, A solution method for optimal weight design problem of the gear using genetic algorithms, Comp. Ind. Eng., 35 (1998), 523-526. http://doi.org/10.1016/S0360-8352(98)00149-1 doi: 10.1016/S0360-8352(98)00149-1
![]() |
[11] |
A. H. Gandomi, X-S. Yang, A. H. Alavi, Cuckoo search algorithm: a metaheuristic approach to solve structural optimization problems, Eng. Comp., 29 (2013), 17-35. http://doi.org/10.1007/s00366-011-0241-y doi: 10.1007/s00366-011-0241-y
![]() |
[12] | E. S. Maputi, R. Arora, Design optimization of a three-stage transmission using advanced optimization techniques, Int. J. Simul. Multidisci. Des. Opt., 10 (2019). http://doi.org/10.1051/smdo/2019009 |
[13] |
M. Castelli, L. Vanneschi, Genetic algorithm with variable neighborhood search for the optimal allocation of goods in shop shelves, Oper. Res. Let., 42 (2014), 355-360. http://doi.org/10.1016/j.orl.2014.06.002 doi: 10.1016/j.orl.2014.06.002
![]() |
[14] |
S. Babaie-Kafaki, R. Ghanbari, N. Mahdavi-Amiri, Hybridizations of genetic algorithms and neighborhood search metaheuristics for fuzzy bus terminal location problems, App. Soft Comp., 46 (2016), 220-229. http://doi.org/10.1016/j.asoc.2016.03.005 doi: 10.1016/j.asoc.2016.03.005
![]() |
[15] |
O. Dib, M-A. Manier, L. Moalic, A. Caminada. Combining VNS with genetic algorithm to solve the one-to-one routing issue in road networks, Comp. Oper. Res., 78 (2017), 420-430. http://doi.org/10.1016/j.cor.2015.11.010 doi: 10.1016/j.cor.2015.11.010
![]() |
[16] |
O. Dib, A. Moalic, M-A. Manier, A. Caminada. An advanced GA-VNS combination for multicriteria route planning in public transit networks, Exp. Sys. with Appl., 72 (2017), 67-82. http://doi.org/10.1016/j.eswa.2016.12.009 doi: 10.1016/j.eswa.2016.12.009
![]() |
[17] |
M. Gen, L. Lin, Y. Yun, H. Inoue, Recent advances in hybrid priority-based genetic algorithms for logistics and SCM network design, Comp. Ind. Eng., 115 (2018), 394-412. http://doi.org/10.1016/j.cie.2018.08.025 doi: 10.1016/j.cie.2018.08.025
![]() |
[18] |
Y. Yun, A. Chuluunsukh, M. Gen, Sustainable closed-loop supply chain design problem: A hybrid genetic algorithm approach, Mathematics, 8 (2020), 84. http://doi.org/10.3390/math8010084 doi: 10.3390/math8010084
![]() |
[19] |
I. Sbai, S. Krichen, O. Limam, Two meta-heuristics for solving the capacitated vehicle routing problem: the case of the Tunisian Post Office, Oper. Res., 20 (2020), 2085-2108. http://doi.org/10.1007/s12351-019-00543-8 doi: 10.1007/s12351-019-00543-8
![]() |
[20] |
J. Wu, M. Fan, Y. Liu, Y. Zhou, N. Yang, M. Yin, A hybrid ant colony algorithm for the winner determination problem, Math. Biosci. Eng., 19 (2022), 3202-3222. http://doi:10.3934/mbe.2022148 doi: 10.3934/mbe.2022148
![]() |
[21] | Y. Yun, C. U. Moon, Comparison of adaptive genetic algorithm for engineering optimization problems, Int. J. Ind. Eng., 10 (2003), 584-590. |
[22] | K. Nitisiri, H. Ohwada, M. Gen, Hybrid genetic algorithm with auto-tuning parameters and K-mean clustering strategy for multiple optimization, J. Soc. Plant Eng. Japan, 31 (2019), 58-67. |
[23] |
Y-T. Kao, E. Zahara, A hybrid genetic algorithm and particle swarm optimization for multimodal functions, Appl. Soft Comp., 8 (2008), 849-857. http://doi.org/10.1016/j.asoc.2007.07.002 doi: 10.1016/j.asoc.2007.07.002
![]() |
[24] |
X. Huang, Z. Guan, L. Yang, An effective hybrid algorithm for multi-objective flexible job-shop scheduling problem, Adv. Mech. Eng., 10 (2018), 1-14. http://doi.org/10.1177/1687814018801442 doi: 10.1177/1687814018801442
![]() |
[25] |
M. Güçyetmez, E. Çam, A new hybrid algorithm with genetic-teaching learning optimization (G-TLBO) technique for optimizing of power flow in wind-thermal power systems, Elect. Eng., 98 (2016), 145-157. http://doi.org/10.1007/s00202-015-0357-y doi: 10.1007/s00202-015-0357-y
![]() |
[26] |
C. J. Shih. Y. C. Yang, Generalized Hopfield network based structural optimization using sequential unconstrained minimization technique with additional penalty strategy, Adv. Eng. Sof., 33 (2002), 721-729. http://doi.org/10.1016/S0965-9978(02)00060-1 doi: 10.1016/S0965-9978(02)00060-1
![]() |
[27] | Y. Yun, Study on adaptive hybrid genetic algorithm and its applications to engineering design problems, Ph.D. Thesis, Waseda University, Japan, 2005. |
[28] | D. Kvalie, Optimization of plane elastic grillages, PhD Thesis, Norges Teknisk Naturvitenskapelige Universitet, Norway, 1967. |
[29] |
T. Ray, P. Saini, Engineering design optimization using a swarm with an intelligent information sharing among individuals, Eng. Opt., 33 (2007), 735-748. http://doi.org/10.1080/03052150108940941 doi: 10.1080/03052150108940941
![]() |
[30] | A. H. Gandomi, X. S. Yang, Benchmark problems in structural optimization, Chapter 12 in Comp. Opt., Meth. and Alg., (eds. S. Koziel, X-S. Yang) Springer-Verlag, Berlin, (2011), 267-291. http://doi.org/10.1007/978-3-642-20859-1_12 |
[31] | M. Gen, R. Cheng, Genetic algorithms and engineering optimization, John Wiley & Sons, New York, NY, USA, 2000. |
[32] | J. Kennedy, R. C. Eberhart, Particle swarm optimization, in Proceedings on IEEE International Conference on Neural Networks, (1995), 1942-1948. http://doi:10.1109/ICNN.1995.488968 |
[33] | X. Yu, M. Gen, Introduction to evolutionary algorithms, Springer, London, UK, 2010. |
[34] | R. V. Rao, Teaching learning based optimization algorithm and its engineering applications, Springer, Switzerland, 2016. http://doi.org/10.1007/978-3-319-22732-0_2 |
[35] | M. Gen, R. Cheng, Genetic algorithms and engineering design, John Wiley and Sons, New York, 1997. |
[36] | Z. Michalewicz, Genetic algorithms + data structures = evolution program, Spring-Verlag, 1994. |
[37] |
Y. Marinakis, M. Marinaki, A hybrid genetic - Particle swarm optimization algorithm for the vehicle routing problem, Exp. Syst. Appl., 37 (2010), 1446-1455. http://doi.org/10.1016/j.eswa.2009.06.085 doi: 10.1016/j.eswa.2009.06.085
![]() |
[38] |
H. Zhai, Y. K. Liu, K. Yang, Modeling two-stage UHL problem with uncertain demands, Appl. Math. Model., 40 (2016), 3029-2048. http://doi.org/10.1016/j.apm.2015.09.086 doi: 10.1016/j.apm.2015.09.086
![]() |
[39] | M. Gen, A. Chuluunsukh, Y. Yun, Hybridizing teaching-learning based optimization with GA and PSO: Case study of supply chain network model, in the 2021 International Conference on Computational Science and Computational Intelligence (CSCI 2021), Las Vegas, USA, (2021). http://doi:10.1109/CSCI54926.2021.00146 |
1. | Amjad Ali, Muhammad Arshad, Eskandar Ameer, Asim Asiri, Certain new iteration of hybrid operators with contractive M -dynamic relations, 2023, 8, 2473-6988, 20576, 10.3934/math.20231049 | |
2. | Muhammad Tariq, Muhammad Arshad, Mujahid Abbas, Eskandar Ameer, Saber Mansour, Hassen Aydi, A relation theoretic m-metric fixed point algorithm and related applications, 2023, 8, 2473-6988, 19504, 10.3934/math.2023995 | |
3. | Imo Kalu Agwu, Naeem Saleem, Umar Isthiaq, A new modified mixed-type Ishikawa iteration scheme with error for common fixed points of enriched strictly pseudocontractive self mappings and ΦΓ-enriched Lipschitzian self mappings in uniformly convex Banach spaces, 2025, 26, 1989-4147, 1, 10.4995/agt.2025.17595 | |
4. | Muhammad Tariq, Sabeur Mansour, Mujahid Abbas, Abdullah Assiry, A Solution to the Non-Cooperative Equilibrium Problem for Two and Three Players Using the Fixed-Point Technique, 2025, 17, 2073-8994, 544, 10.3390/sym17040544 |