Hybrid decision support system disaster management: application of lattice ordered q-rung linear Diophantine fuzzy hypersoft sets

View article
PeerJ Computer Science

Introduction

The frequent occurrence of uncertainty-related issues in multi-attribute decision-making (MADM) makes them difficult to foresee and manage due to the extensive modeling of these uncertainties. The fuzzy set (FS) theory introduced by Zadeh (1965) is very useful for handling the difficulties brought on by uncertainty. However, FS only has a limited ability to reflect impartial situations. To overcome these restrictions, Atanassov (1986) devised the notion of intuitionistic fuzzy sets (IFS). The IFS’s two indices are membership degree (MD) and non-membership degree (NMD), and their sum value should fall within [0,1]. To solve problems smoothly, Yager (2013) developed the Pythagorean fuzzy set (PFS) in which the total of the MD2 and NMD2 should fall within [0,1]. Yager (2016) also proposed the q-rung orthopair fuzzy sets (q-ROFS), where the MDq and the NMDq are summed together and fall inside the range [0,1]. Later, various information measures (Peng & Liu, 2019) were proposed for q-ROFS. However, each of these ideas has drawbacks of its own. To overcome these drawbacks, Riaz & Hashmi (2019) formulated the theory of the linear Diophantine fuzzy set (LDFS), which contains the notion of reference parameters (RPs). Owing to the usefulness of LDFS, several researchers from various scientific fields were interested in them, and numerous significant studies were produced as a result (Mahmood et al., 2021a, 2021b). Subsequently, the idea of quadratic diophantine fuzzy set was proposed by Zia et al. (2023). Later, Almagrabi et al. (2022) created the q-rung linear Diophantine fuzzy set (q-RLDFS), a particular extension of the IFS, q-ROFS, and LDFS. Further, many real-world decision-making studies such as company selection problem (Ali, 2025), urban planning (Petchimuthu et al., 2025), logistics (Kannan, Jayakumar & Pethaperumal, 2025) and emerging technologies (Kumar & Pamucar, 2025). However, because they are not parametrized, each theory has drawbacks. To overcome the limitations brought on by parametrization, Molodtsov (1999) developed the idea of soft set (SS) theory, which handles vagueness in a parametric manner. Later, by incorporating FS and SS, Roy & Maji (2007) provided the idea of the fuzzy soft set (FSS), which helps present fuzzy data with parametric information. Similar to this, SS theory was incorporated with other extensions of FS theory such as IFS, PFS, q-ROFS, and LDFS (Çağman & Karataş, 2013; Peng et al., 2015; Hussain et al., 2020; Riaz et al., 2020) respectively, to exhibit these fuzzy extension data with parametric information and obtained intuitionistic fuzzy soft set (IFSS), Pythagorean fuzzy soft set (PFSS), q-rung orthopair fuzzy soft set (q-ROFSS) and linear Diophantine fuzzy soft set (LDFSS). Smarandache (2018) then transformed the function into a multi-attributed function to establish the idea of the hypersoft set (HSS) as an extension of SS. By incorporating HSS with FS and IFS, Smarandache (2018) also proposed the ideas of the fuzzy hypersoft set (FHSS) and intuitionistic fuzzy hypersoft set (IFHSS), which expresses FS and IFS data with multi-sub-parameter. Similarly, by incorporating q-ROFS with HSS, Khan, Gulistan & Wahab (2022) presented the q-rung orthopair fuzzy hypersoft set (q-ROFHSS), and by incorporating q-RLDFS with HSS, Surya et al. (2024) presented the q-rung linear Diophantine fuzzy hypersoft set (q-RLDFHSS). In many real-life problems, there is a ranking among the parameters to deal with such problems very effectively. Ali et al. (2015) proposed a lattice-ordered soft set (LOSS). Later, Aslam et al. (2019) discussed the notion of lattice-ordered fuzzy soft set (LOFSS), and Mahmood et al. (2018) discussed the notion of lattice-ordered intuitionistic fuzzy soft set (LOIFSS). Further, many researchers (Rajareega & Vimala, 2021; Pandipriya, Vimala & Begam, 2018; Mahmood, Rehman & Sezgin, 2018; Begam et al., 2020; Khan, Bakhat & Iftikhar, 2019; Sabeena Begam & Vimala, 2019) developed the concepts of lattice-ordered structure to various areas of FS theory and their extensions. Likewise, to discuss real-life q-RLDFHS problems when there is a ranking among the multi-sub-parameters the notion of lattice ordered q-rung linear Diophantine fuzzy hypersoft set (LOq-RLDFHSS) is essential.

Research gap

Listed below are the research gaps:

  • From the analysis of existing literature, we can see that in theoretical aspects, the existing literature does not cover many fundamental algebraic operations of LOq-RLDFHSS.

  • Further from the existing literature, we can see that while there are several parametric decision-making (DM) studies conducted under various fuzzy structures, it is challenging to demonstrate many MADM real-world problems under LOq-RLDFHS environment using the existing literature.

Motivation

The following are the study’s motivations:

  • The study aims to close these research gaps by developing fundamental algebraic operations and a MADM method based on LOq-RLDFHSS.

  • Another main motive of the study is to contribute to the disaster management field by the proposed MADM approach, since the existing DM methods in the disaster management field cannot handle multiple attributes simultaneously.

Objectives

The main objectives of this work are listed below:

  • To provide many fundamental algebraic operations of LOq-RLDFHSS.

  • To provide an effective MADM strategy based on LOq-RLDFHSS.

  • To provide an appropriate numerical illustration for the suggested MADM strategy in the field of disaster management.

Contribution

The core contributions of the work are as follows:

  • Many algebraic operators of LOq-RLDFHSS are proposed in this study, such as restricted union, restricted intersection, extended union, OR operation, AND operation, and complement.

  • A MADM algorithm based on the LOq-RLDFHSS is presented in the study.

  • Further, a real-world problem in the field of disaster management is depicted as a numerical example of the suggested MADM algorithm to show the efficacy of the proposed algorithm.

  • To demonstrate the potency and efficacy of the suggested concepts and the MADM approach, a comparative assessment that describes the theoretical improvement of the proposed study and its contribution to the field of disaster management is presented, along with the minor restrictions of the proposed concepts.

The list of most of the abbreviations used in this study is given as a table in “List of abbreviation used in the study”. The article is structured as follows:

“Background” contains the required introductory notations and definitions. “Algebraic operations of LOq-RLDFHSS” consists of fundamental algebraic operations of LOq-RLDFHSS. “MADM Approach Based on LOq-RLDFHSS” consists of a MADM algorithm based on LOq-RLDFHSS to successfully solve MADM challenges; a MADM problem in disaster management which demonstrates the efficiency of the proposed algorithm. To describe the superiority of the proposed idea to the existing ideas, a comparative assessment has been undertaken in “Comparative Assessment”. Finally, “Conclusion” provides the conclusion of the article.

Background

This section provides the requisite notations and definitions for this article.

A binary relation on a non-empty set A is said to be a partial order on A if it is antisymmetric, reflexive and transitive. Also, is said to be a total order on A if ab, either ab or ba a,bA.

A partial order set L is said to be a lattice if the set {a,b} has a greatest lower bound and least upper bound a,b L. If L contains 1 and 0 such that x L, 0 x 1, then L is called a bounded lattice.

Definition 2.1. Atanassov (1986): Let G be the set of alternatives. A IFS I is defined as

I={(g,ΩI(g),I(g))|gG}where ΩI(g) and I(g) [0,1] are MD and NMD fulfilling 0 ΩI(g)+I(g) 1.

Definition 2.2. Almagrabi et al. (2022): Let G be the set of alternatives. A q-RLDFS Q is defined as

Q={(g,ΩQ(g),Q(g),ΔQ(g),Q(g))|gG}where ΩQ(g),Q(g),ΔQ(g) and Q(g) [0,1] are MD, NMD and their corresponding RPs respectively, fulfilling 0 ΔQq(g)+Qq(g) 1 and 0 ΔQq(g)ΩQ(g)+Qq(g)Q(g) 1 gG, q 1.

Definition 2.3. Molodtsov (1999): Let G be the set of alternatives, E be the set of attributes, and AE. Then SS is a pair (Θ,A) defined by the mapping

Θ:AP(G)where P (G) is the power set of G.

Definition 2.4. Ali et al. (2015): Let (Θ,A) be a SS over G, where

Θ:AP(G)Then (Θ,A) is said to be a LOSS if a1  Aa2Θ(a1)Θ(a2)a1,a2A.

Definition 2.5. Çağman & Karataş (2013): Let G be the set of alternatives, E be the set of attributes, and AE. Then IFSS is a pair (Θ,A) defined by the mapping

Θ:AIFP(G)where IFP (G) is the IF power set of G.

Definition 2.6. Mahmood et al. (2018): Let (Θ,A) be a IFSS over G, where

Θ:AIFP(G)Then (Θ,A) is said to be a LOIFSS if a1 Aa2Θ(a1)Θ(a2)a1,a2A.

Definition 2.7. Smarandache (2018): Let G be the set of alternatives and P(G) denote the Power set of G. Let E1,E2,...,En with EiEj= for i,j{1,2,n} and ij be the attribute values of n distinct attributes e1,e2,,en respectively and for each i=1,2,n, Ai be non empty subset of Ei and 1=A1×A2××AnE1×E2××En. Then HSS over G is the pair (Θ,1) defined by the map

Θ:1P(G)This can be represented as (Θ,1)={(η,Θ(η)):η1,Θ(η)P(G)}.

Definition 2.8. Surya et al. (2024): Let G be the set of alternatives and q-RLDFP ( G) denote the q-RLDF Power set of G. Let E1,E2,,En with EiEj= for i,j{1,2,n} and ij be the attribute values of n distinct attributes e1,e2,,en respectively and for each i=1,2,n, Ai be non empty subset of Ei and 1=A1×A2××AnE1×E2××En. Then, the q-Rung Linear Diophantine Fuzzy Hypersoft Set over G (q-RLDFHSS ( G)) is the pair (Θ,1) defined by the map

Θ:1qRLDFP(G)This can be represented as (Θ,1)={(η,Θ(η)):η1,Θ(η)qRLDFP(G)} and the q-RLDFHS Number (q-RLDFHSN)

Θga(ηc) = {ΩΘ(ηc)(ga),Θ(ηc)(ga),ΔΘ(ηc)(ga),Θ(ηc)(ga)|gaGandηc1} can be express as Jηac = {Ωηac,ηac,Δηac,ηac}.

Definition 2.9. Surya et al. (2024): Let (Θ1,1),(Θ2,2) q-RLDFHSS ( G), then (Θ1,1) is said to be q-RLDFHS subset of (Θ2,2), if

(i) 12

(ii) η1,Θ1(η)Θ2(η)

(i.e.,) ΩΘ1(η)(ga)ΩΘ2(η)(ga),Θ2(η)(ga)Θ1(η)(ga),ΔΘ1(η)(ga)ΔΘ2(η)(ga) and Θ2(η)(ga)Θ1(η)(ga)gaG.

Algebraic operations of loq-rldfhss

In this section, the fundamental algebraic operations of LOq-RLDFHSS are presented.

Definition 3.1. A q-RLDFHSS (G) (Θ,A1×A2×...×An=1) is said to be lattice ordered q-RLDFHSS over G (LOq-RLDFHSS ( G)) if for mapping Θ:1q-RLDFP(G),

η11η2Θ(η1)Θ(η2) η1,η21

(i.e.,) η11η2

   ΩΘ(η1)(ga)ΩΘ(η2)(ga), Θ(η2)(ga)Θ(η1)(ga),

    ΔΘ(η1)(ga)ΔΘ(η2)(ga) and Θ(η2)(ga)Θ(η1)(ga) gaG

where η1=(η11,η12,,η1n),η2=(η21,η22,,η2n) and η1i,η2iAi for i{1,2,,n}.

Also, each Ai is defined by its corresponding binary relation Ai and 1 forms a relation defined by (η11,η12,,η1n)1(η21,η22,,η2n)η1iAi η2i in Ai for i{1,2,,n}.

The following example clarifies the definition above.

EXAMPLE 1. Let G={g1,g2,g3} be the set of hotels for accommodation, consider the attributes e1 = {charges}, e2={food}, e3={service} and E1 = {extra charges (e11), room rent (e12)}, E2 = {taste (e21), hygiene (e22)}, E3 = {customer service (e31)} be their corresponding attribute values respectively.

Suppose that,

For each i = 1, 2, 3, Ai = Ei

The elements in each set A1,A2andA3 have an order among them, they are

The elements in set A1 are in the order e11 A1 e12

The elements in set A2 are in the order e21 A2 e22

A3 has only one element e31 and

=A1×A2×A3={η1=(e11,e21,e31),η2=(e11,e22,e31),η3=(e12,e21,e31),

η4=(e12,e22,e31)}

Then the order of elements in set is shown in Fig. 1.

The order among elements in 
$\aleph$ℵ
.

Figure 1: The order among elements in .

Further, the following is how the attributes are categorized

  • The attribute “charges” and its attribute values indicates whether the alternative is cheap or not cheap

  • The attribute “food” and its attribute values indicates whether the alternative is good or not good

  • The attribute “service” and its attribute values indicates whether the alternative satisfies or dissatisfies

Then, the Cartesian product of attribute values exemplifies that the alternative is (cheap, good, satisfies) altogether or (not cheap, not good, dissatisfies) altogether.

Then, q-RLDFHSS ( Θ,) may be expressed as

(Θ,)={η1,(g1(0.4,0.8),(0.3,0.9),g2(0.3,0.7),(0.4,0.9),g3(0.4,0.7),(0.2,0.7)),η2,(g1(0.4,0.7),(0.4,0.8),g2(0.4,0.6),(0.5,0.7),g3(0.5,0.6),(0.4,0.6)),η3,(g1(0.5,0.7),(0.5,0.8),g2(0.4,0.6),(0.5,0.8),g3(0.4,0.6),(0.3,0.6)),η4,(g1(0.6,0.6),(0.5,0.6),g2(0.7,0.6),(0.7,0.5),g3(0.8,0.4),(0.7,0.3))}

We will assume that q = 3.

The characteristic of this q-RLDFHSS (Θ,) is (MD, NMD ,(cheap, good, satisfies), (not cheap, not good, dissatisfies) ) ηc.

Clearly Θ(η1)Θ(η2)Θ(η4) and Θ(η1)Θ(η3)Θ(η4), therefore, (Θ,1) is a LOq-RLDFHSS (G).

Definition 3.2. Let G be the set of alternatives and (Θ1,1),(Θ2,2) LOq-RLDFHSS (G). Their Restricted union is defined by (Θ1,1) RES (Θ2,2)=(Θ3,3) where 3=12 and η3,gG we have Θ1(η)Θ2(η)=Θ3(η).

ΩΘ3(η)(g)=Max{ΩΘ1(η)(g),ΩΘ2(η)(g)},

Θ3(η)(g)=Min{Θ1(η)(g),Θ2(η)(g)},

ΔΘ3(η)(g)=Max{ΔΘ1(η)(g),ΔΘ2(η)(g)} and

Θ3(η)(g)=Min{Θ1(η)(g),Θ2(η)(g)}.

Proposition 3.3. Let (Θ1,1),(Θ2,2) LOq-RLDFHSS (G). Then (Θ1,1) RES (Θ2,2) LOq-RLDFHSS (G).

Proof. See “Proof of Proposition 3.2”. □

Definition 3.4. Let G be the set of alternatives and (Θ1,1),(Θ2,2) LOq-RLDFHSS (G). Their Restricted intersection is defined by (Θ1,1)RES(Θ2,2)=(Θ3,3) where 3=12 and η3,gG we have Θ1(η)Θ2(η)=Θ3(η).

ΩΘ3(η)(g)=Min{ΩΘ1(η)(g),ΩΘ2(η)(g)},

Θ3(η)(g)=Max{Θ1(η)(g),Θ2(η)(g)},

ΔΘ3(η)(g)=Min{ΔΘ1(η)(g),ΔΘ2(η)(g)} and

Θ3(η)(g)=Max{Θ1(η)(g),Θ2(η)(g)}.

Proposition 3.5. Let (Θ1,1),(Θ2,2) LOq-RLDFHSS (G). Then (Θ1,1)RES(Θ2,2) LOq-RLDFHSS (G).

Proof. See “Proof of Proposition 3.4”. □

Definition 3.6. Let G be the set of alternatives and (Θ1,1),(Θ2,2) LOq-RLDFHSS (G). Their extended union is defined by (Θ1,1) EXT(Θ2,2)=(Θ3,3) where 3=12

(Θ3,3)={{ΩΘ1(η)(g),Θ1(η)(g),ΔΘ1(η)(g),Θ1(η)(g)}ifη12{ΩΘ2(η)(g),Θ2(η)(g),ΔΘ2(η)(g),Θ2(η)(g)}ifη21{Max{ΩΘ1(η)(g),ΩΘ2(η)(g)},Min{Θ1(η)(g),Θ2(η)(g)},ifη12Max{ΔΘ1(η)(g),ΔΘ2(η)(g)},Min{Θ1(η)(g),Θ2(η)(g)}}

Proposition 3.7. Let (Θ1,1),(Θ2,2) LOq-RLDFHSS (G). Then (Θ1,1) EXT(Θ2,2) LOq-RLDFHSS (G), if one of them is a LOq-RLDFHSS subset of other.

Proof. See “Proof of Proposition 3.6”. □

Definition 3.8. Let 1,2E1×E2××En. Then partial order 1×2 on 1×2 is defined as for any (η1,ς1),(η2,ς2)1×2,(η1,ς1)1×2(η2,ς2) η1 1 η2 and ς12 ς2.

Definition 3.9. Let G be the set of alternatives and (Θ1,1),(Θ2,2) LOq-RLDFHSS (G). Their “AND” operation is defined by (Θ1,1)(Θ2,2)=(Ξ,1×2)

where

Ξ(1×2)={(η,ς),(g,Ξ(η,ς)(g)):gG,(η,ς)1×2}and Ξ(η,ς)(g)={Min{ΩΘ1(η)(g),ΩΘ2(ς)(g)},Max{Θ1(η)(g),Θ2(ς)(g)},

Min{ΔΘ1(η)(g),ΔΘ2(ς)(g)},Max{Θ1(η)(g),Θ2(ς)(g)}}.

Proposition 3.10. Let G be the set of alternatives and (Θ1,1),(Θ2,2) LOq-RLDFHSS (G). Then (Θ1,1)(Θ2,2) LOq-RLDFHSS (G).

Proof. See “Proof of Proposition 3.9”. □

Definition 3.11. Let (G) be the set of alternatives and (Θ1,1),(Θ2,2) LOq-RLDFHSS (G). Then their “OR” operation is defined by (Θ1,1)(Θ2,2)=(Ξ,1×2)

where

(Ξ,1×2)={(η,ς),(g,Ξ(η,ς)(g)):gG,(η,ς)1×2} and Ξ(η,ς)(g)={Max{ΩΘ1(η)(g),ΩΘ2(ς)(g)},Min{Θ1(η)(g),Θ2(ς)(g)},

Max{ΔΘ1(η)(g),ΔΘ2(ς)(g)},Min{Θ1(η)(g),Θ2(ς)(g)}}

Proposition 3.12. Let G be the set of alternatives and (Θ1,1),(Θ2,2) LOq-RLDFHSS (G). Then (Θ1,1)(Θ2,2) LOq-RLDFHSS (G).

Proof. See “Proof of Proposition 3.11”. □

Definition 3.13. Let (Θ1,1) LOq-RLDFHSS (G).

If ΩΘ1(η)(g) = ΔΘ1(η)(g) = 0, Θ1(η)(g) = Ω1(η)(g) = 1 η1 and gG, Then, (Ω1,1) is called the relative null LOq-RLDFHSS and is denoted by 1.

Definition 3.14. Let (Θ1,1) LOq-RLDFHSS (G).

If ΩΘ1(η)(g) = ΔΘ1(η)(g) = 1, Θ1(η)(g) = Ω1(η)(g) = 0 η1 and gG, Then, (Θ1,1) is called the relative universal LOq-RLDFHSS and is denoted by U1.

Proposition 3.15. Let (Θ1,1) LOq-RLDFHSS (G). Then

1. (Θ1,1)RES(Θ1,1) = (Θ1,1)

2. (Θ1,1)RES1 = (Θ1,1)

3. (Θ1,1)RESU1 = U1

4. (Θ1,1)RES(Θ1,1) = (Θ1,1)

5. (Θ1,1)RES1 = 1

6. (Θ1,1)RESU1 = (Θ1,1)

Proof. Straightforward. □

Definition 3.16. Let (Θ1,1) LOq-RLDFHSS (G). Then complement of (Θ1,1) denoted by (Θ1,1)c and is defined as follows

(Θ1,1)c = {(g,{Θ1(η)(g),ΩΘ1(η)(g),Θ1(η)(g),ΔΘ1(η)(g)}):η1 and gG}.

Proposition 3.17. Let (Θ1,1) LOq-RLDFHSS (G). Then ((Θ1,1)c)c=(Θ1,1)

Proof. Let (Θ1,1) LOq-RLDFHSS (G). Then complement of (Θ1,1) is

(Θ1,1)c = {(g,{Θ1(η)(g),ΩΘ1(η)(g),Θ1(η)(g),ΔΘ1(η)(g)}):η1 and gG},

Now complement of (Θ1,1)c is

((Θ1,1)c)c = {(g,{ΩΘ1(η)(g),Θ1(η)(g),ΔΘ1(η)(g),Θ1(η)(g)}):η1 and gG} = (Θ1,1). □

EXAMPLE 2. Let G={g1,g2} be a set of alternatives, 1={η1,η2} be a set of parameters with an order defined by η1 η2 and 2={η1,η3} be another set of parameters with an order defined by η1 η3. Then, let

(Θ1,1)={η1,(g1(0.3,0.8),(0.4,0.8),g2(0.2,0.8),(0.2,0.8)),η2,(g1(0.6,0.5),(0.5,0.7),g2(0.6,0.5),(0.6,0.4))}be a q-RLDFHSS with q as 3, and since Θ1(η1)Θ1(η2), this implies (Θ1,1) is a LOq-RLDFHSS. Also, let

(Θ2,2)={η1,(g1(0.5,0.8),(0.4,0.7),g2(0.3,0.7),(0.2,0.8)),η3,(g1(0.6,0.7),(0.6,0.6),g2(0.7,0.3),(0.9,0.2))}be another q-RLDFHSS with q as 3, and since Θ2(η1)Θ2(η3), this implies (Θ2,2) is a LOq-RLDFHSS.

The following operations are then derived:

  • (Θ1,1)RES(Θ2,2)={η1,(g1(0.5,0.8),(0.4,0.7),g2(0.3,0.7),(0.2,0.8))}

  • (Θ1,1)RES(Θ2,2)={η1,(g1(0.3,0.8),(0.4,0.8),g2(0.2,0.8),(0.2,0.8))}

  • (Θ1,1)EXT(Θ2,2)={η1,(g1(0.5,0.8),(0.4,0.7),g2(0.3,0.7),(0.2,0.8)),η2,(g1(0.6,0.5),(0.5,0.7),g2(0.6,0.5),(0.6,0.4)),η3,(g1(0.6,0.7),(0.6,0.6),g2(0.7,0.3),(0.9,0.2))}

  • (Θ1,1)(Θ2,2)={(η1,η1),(g1(0.5,0.8),(0.4,0.7),g2(0.3,0.7),(0.2,0.8)),(η1,η3),(g1(0.6,0.7),(0.6,0.7),g2(0.7,0.3),(0.9,0.2)),(η2,η1),(g1(0.6,0.5),(0.5,0.7),g2(0.6,0.5),(0.6,0.4)),(η2,η3),(g1(0.6,0.5),(0.6,0.7),g2(0.7,0.3),(0.9,0.2))}

  • (Θ1,1)(Θ2,2)={(η1,η1),(g1(0.3,0.8),(0.4,0.8),g2(0.2,0.8),(0.2,0.8)),(η1,η3),(g1(0.3,0.8),(0.4,0.8),g2(0.2,0.8),(0.2,0.8)),(η2,η1),(g1(0.5,0.8),(0.4,0.7),g2(0.3,0.7),(0.2,0.8)),(η2,η3),(g1(0.6,0.7),(0.5,0.7),g2(0.6,0.5),(0.6,0.4))}

Madm approach based on loq-rldfhss

In this section, the comparison matrix of LOq-RLDFHSS and a MADM algorithm based on LOq-RLDFHSS are described and a MADM problem in the field of disaster management is discussed as a numerical illustration for the proposed MADM algorithm.

Definition 4.1. The comparison matrix of LOq-RLDFHSS is a matrix in which rows represent the alternatives such as g1,g2,,gm and columns represent the parameters η1,η2,,ηr. The entries are hac and computed as hac=k1k2+k3k42, where k1 is the integer computed as number of times ΩΘ(ηc)(ga) greater than or equal to ΩΘ(ηc)(gb), for gagb,gbG,k2 is the integer computed as number of times Θ(ηc)(ga) greater than or equal to Θ(ηc)(gb), for gagb,gbG,k3 is the integer computed as number of times ΔΘ(ηc)(ga) greater than or equal to ΔΘ(ηc)(gb), for gagb,gbG and k4 is the integer computed as number of times Θ(ηc)(ga) greater than or equal to Θ(ηc)(gb), for gagb,gbG. Further, the range of hac lies within [(m1),m1].

Definition 4.2. The score of an alternative ga is Sa and calculated as

Sa=c=1rhacwhere the range lies within [r((m1)),r(m1)].

Algorithm

The following steps describe the algorithm for selecting the most suitable alternative

Step 1: Consider the LOq-RLDFHSS ( G) and keep it in tabular form

Step 2: Compute the comparison matrix of LOq-RLDFHSS.

Step 3: Calculate the score Sa of ga a

Step 4: Find Sl = Max Sa and choose it as the suitable alternative

Step 5: If multiple alternatives share the maximum score, select any one of them.

Figure 2 shows the proposed algorithm as a flowchart.

Flowchart showing the steps of the proposed LOq-RLDFHSS-based MADM algorithm.

Figure 2: Flowchart showing the steps of the proposed LOq-RLDFHSS-based MADM algorithm.

Numerical illustration

A general study about disaster management

Disaster management or emergency management is the administrative responsibility for creating the framework that assists societies in reducing their vulnerability to hazards and coping with calamities. Contrary to its name, disaster management does not focus on handling crises, which are typically regarded as minor occurrences with little consequences that are dealt with through regular community activities. The main goal of emergency management is the management of disasters, which are occurrences with more consequences than a community can manage on its own. A mix of efforts by individuals, households, businesses, local governments, and/or higher levels of government is typically required for disaster management. Even though the discipline of emergency management uses a variety of terminologies, operations can generally be broken down into four categories: preparedness, response, mitigation, and recovery. In other words, mitigation of disaster risks and prevention are also frequently used.

The guiding principle of disaster management is disaster mitigation. The continuous work aims to reduce disasters’ harm to both persons and property. Mitigation measures include avoiding constructing near floodplains, designing bridges to resist earthquakes, developing and enforcing hurricane-proof building regulations, and more. Mitigation refers to sustained actions that minimize or prevent long-term danger to individuals and assets from environmental risks and their effects.” Disaster consequences are continuously being lessened by federal, state, municipal, and individual actions.

Authorities and organizations on a national or international scale may provide this assistance during disaster. Effective coordination of disaster assistance is frequently crucial when numerous organizations contribute to the response, but competence has been degraded by the disaster or overwhelmed by demand. The US government released an article called the National Response Framework (Federal Emergency Management Agency, 2023) that outlines the responsibilities of authorities of the state, local, national, and tribal governments. It offers guidance on how to fully or partially implement disaster support services to aid in the response and recovery process.

The recovery phase begins once there is no immediate danger to human life. Getting the afflicted area back to normal as soon as possible is the urgent goal of the recovery phase. Trained laypeople give psychological first aid in the early wake of a disaster to help the affected populace cope and recover. In addition to providing practical support and assisting with procuring necessities like food and water, trained staff can also provide links to important resources. Similar to medical first aid, psychological first aid does not require therapists to be licensed clinicians. It is not debriefing, counseling, or psychotherapy.

Numerous research such as disaster management cycle of natural disaster (Arifah, Tariq & Juni, 2019), large scale group decision making in disaster management (Wan et al., 2020), post-disaster reconstruction projects (Mohammadnazari et al., 2022), use of indicators in vulnerability assessment (Papathoma-Köhle et al., 2019) in decision-making have been carried out in disaster management. Now, we show the utilization of proposed conceptions and algorithms in real life by a MADM problem in the field of disaster management, which helps to choose the most appropriate plan to tackle the known upcoming natural disaster by considering more attributes together. The problem is presented below, and its contribution to the disaster management field is discussed in detail in the comparative assessment section.

Problem

Suppose disaster management wants to choose the most appropriate plan from a set of plans {g1, g2, g3} to tackle some of the known upcoming natural disasters as a precautionary measure and a team of decision makers was appointed to analyze the plans, the decision makers are considering the attributes e1 = {mitigation}, e2={response}, e3={recovery} and their sub attributes are E1={education and awareness programs (e11), regulation and infrastructure projects (e12)}, E2={maintaining regular services and activities (e21), protecting life (e22)} and E3={psychological recovery (e31)} respectively. Also, the order of preference of elements in each set E1, E2 and E3 by decision-makers is given as follows

The elements in set E1 are in the order e11E1 e12

The elements in set E2 are in the order e21E2 e22

E3 has only one element e31 and 1=E1×E2×E3={η1=(e11,e21,e31),η2=(e11,e22,e31),η3=(e12,e21,e31), η4=(e12,e22,e31)}

Then the order of elements in set 1 is shown in the Fig. 3.

The order among elements in 
${\aleph_{1}}$ℵ1
.
Figure 3: The order among elements in 1.

Further, decision-makers categorize the attributes as follows:

  • The attribute “mitigation” and its attribute values indicates whether the plan is high or low

  • The attribute “response” and its attribute values indicates whether the plan is good or not good

  • The attribute “recovery” and its attribute values indicates whether the plan is effective or not effective

Then, the Cartesian product of attribute values exemplifies that the plan is (high, good, effective) all together or (low, not good, not effective) all together.

The opinions and data observed by the decision makers are constructed and expressed as a q-RLDFHSS (Θ,1).

The characteristic of this q-RLDFHSS (Θ,1) is (MD, NMD ,(high, good, effective), (low, not good, not effective) ) ηc1.

(Θ,1)={η1,(g1(0.33,0.87),(0.31,0.82),g2(0.29,0.76),(0.33,0.81),g3(0.38,0.63),(0.17,0.72)),η2,(g1(0.4,0.65),(0.38,0.71),g2(0.32,0.57),(0.43,0.66),g3(0.53,0.61),(0.39,0.51)),η3,(g1(0.55,0.66),(0.57,0.72),g2(0.35,0.53),(0.52,0.81),g3(0.46,0.54),(0.24,0.57)),η4,(g1(0.63,0.58),(0.65,0.69),g2(0.63,0.49),(0.74,0.48),g3(0.83,0.41),(0.72,0.28))}

We will assume that q = 3.

Clearly Θ(η1)Θ(η2)Θ(η4) and Θ(η1)Θ(η3)Θ(η4), therefore, (Θ,1) is a LOq-RLDFHSS (G).

In this LOq-RLDFHSS, the plan g1 and the parameter η1 = (education and awareness programs, maintaining regular services and activities, psychological recovery) has the numeric value (0.33,0.87),(0.31,0.82). This value expresses that for the parameter η1 the plan g1 has 33 % truth value and 87% false value. The pair (0.31,0.82) indicates the RP of the truth and false values, respectively, where we can observe that for (high at education and awareness programs, good at maintaining regular services and activities, effective in psychological recovery) all together the plan g1 expresses 31% and for (low at education and awareness programs, not good at maintaining regular services and activities, not effective in psychological recovery) all together the plan g1 expresses 82%. Similarly, all other numeric values are expressed in this LOq-RLDFHSS.

Step 1: Tabular form of LOq-RLDFHSS (Θ,1) is shown in Table 1.

Table 1:
Tabular form of LOq-RLDFHSS ( Θ,1) which describes the data observed by the decision makers about the plans according to the parameters.
( Θ,1) g1 g2 g3
η1 (0.33, 0.87), (0.31, 0.82) (0.29, 0.76), (0.33, 0.81) (0.38, 0.63), (0.17, 0.72)
η2 (0.4, 0.65), (0.38, 0.71) (0.32, 0.57), (0.43, 0.66) (0.53, 0.61), (0.39, 0.51)
η3 (0.55, 0.66), (0.57, 0.72) (0.35, 0.53), (0.52, 0.81) (0.46, 0.54), (0.24, 0.57)
η4 (0.63, 0.58), (0.65, 0.69) (0.63, 0.49), (0.74, 0.48) (0.83, 0.41), (0.72, 0.28)
DOI: 10.7717/peerj-cs.2927/table-1

Step 2: Comparision matrix of LOq-RLDFHSS is shown in Table 2.

Table 2:
Comparison matrix of LOq-RLDFHSS ( Θ,1).
( Θ,1) η1 η2 η3 η4
g1 −1 32 12 32
g2 0 12 12 12
g3 1 1 0 32
DOI: 10.7717/peerj-cs.2927/table-2

Step 3: The scores of the alternatives are shown in Table 3.

Table 3:
Score value of alternatives using the comparison matrix described in Table 2.
( Θ,1) g1 g2 g3
Score 72 12 72
DOI: 10.7717/peerj-cs.2927/table-3

From the obtained scores, we observed that g3 is the most appropriate plan to tackle the disaster, and we got the ranking of plans as g1<g2<g3.

Comparative assessment

Validity test

The effectiveness of a MADM strategy depends on the coherence of the qualities, the relationship between the alternatives, and the decision-maker’s objective evaluations. Wang & Triantaphyllou (2008) created three effective validity test criteria, which must be completed for a MADM approach to be considered legitimate.

Test criteria 1: The optimal choice remains the same if one selects a non-ideal alternative over a non-optimal one without changing the weight of any attribute.

Test criteria 2: The transitive nature is necessary for a decision-making approach to be effective.

Test criteria 3: If the decision-making problem is broken down into smaller subproblems, the smaller subproblem’s order has to correspond with the original problem’s order.

An examination of the suggested method’s validity is provided below:

Test criteria 1: Consider the same disaster management problem by replacing the non-ideal alternative g1 with a worse alternative g^, whose numeric values according to the parameters are

{g^,(η1(0.30,0.89),(0.25,0.91),η2(0.35,0.70),(0.33,0.73),η3(0.51,0.68),(0.54,0.76),η4(0.59,0.62),(0.62,0.72))}Then, after analyzing these three alternatives g^, g2, g3 by the proposed method, we obtain ranking as g^<g2<g3. The result makes it clear that the best solution remains constant. Therefore, test criteria 1 is satisfied for the proposed methodology.

Test criteria 2 and 3: We divide the considered problem into sub-problems as {g1,g3},{g1,g2} and {g2,g3}. Then using the proposed method we obtain g1<g3, g1<g2 and g2<g3 as the ranking of sub-problems respectively. Therefore, we can see that the overall ranking remains constant as g1<g2<g3. For the suggested approach, test criteria 2 and 3 are therefore valid.

Comparative analysis

To analyze the superiority of the proposed DM method, the advantages and restrictions of existing and proposed DM methods are described in Table 4.

Table 4:
Comparison table which describes the advantages and restrictions of existing and proposed decision making methods.
DM methods Advantages Restrictions
FS (Zadeh, 1965) Addresses uncertainty by Ω (MD) Unable to deal with (NMD) and parametrization
IFS (Atanassov, 1986) Addresses uncertainty by Ω and Restricted in handling uncertainty by the condition Ω+ [0,1], also unable to deal with parametrization
PFS (Yager, 2013) Addresses uncertainty by Ω and even if Ω+ [0,1] Restricted in handling uncertainty by the condition Ω2+2 [0,1], also unable to deal with parametrization
q-ROFS (Yager, 2016) Addresses uncertainty by Ω and even if Ω2+2 [0,1] Restricted in handling uncertainty by the condition Ωq+q [0,1], also unable to deal with parametrization
LDFS (Riaz & Hashmi, 2019) Addresses uncertainty by Ω, , Δ (RP corresponding to MD) and (RP corresponding to NMD) even if Ωq+q [0,1] Restricted in handling uncertainty by the conditions ΔΩ+ [0,1] and Δ+ [0,1], also unable to deal with parametrization
q-RLDFS (Almagrabi et al., 2022) Addresses uncertainty by Ω, , Δ and even if ΔΩ+ [0,1] and Δ+ [0,1] Restricted in handling uncertainty by the conditions ΔqΩ+q [0,1] and Δq+q [0,1] also unable to deal with parametrization
SS (Molodtsov, 1999) Able to deal with parametrization Unable to address uncertainty by parameterization
FSS (Roy & Maji, 2007) Addresses FS with parameterized values Unable to address uncertainty exceeding FS’s restriction by parameterized values and also unable to address FS by multi-sub-parameterized values
IFSS (Çağman & Karataş, 2013) Addresses IFS with parameterized values Unable to address uncertainty exceeding IFS’s restriction by parameterized values and also unable to address IFS by multi-sub-parameterized values
q-ROFSS (Hussain et al., 2020) Addresses q-ROFS with parameterized values Unable to address uncertainty exceeding q-ROFS’s restriction by parameterized values and also unable to address q-ROFS by multi-sub-parameterized values
LDFSS (Riaz et al., 2020) Addresses LDFS with parameterized values Unable to address uncertainty exceeding LDFS’s restriction by parameterized values and also unable to address LDFS by multi-sub-parameterized values
LOSS (Ali et al., 2015) Addresses SS effectively when there is a ranking among parameters Unable to address uncertainty by parameterization
LOFSS (Aslam et al., 2019) Addresses FSS effectively when there is a ranking among parameters Unable to address uncertainty exceeding FS’s restriction by parameterized values and also unable to address FS by multi-sub-parameterized values
LOIFSS (Mahmood et al., 2018) Addresses IFSS effectively when there is a ranking among parameters Unable to address uncertainty exceeding IFS’s restriction by parameterized values and also unable to address IFS by multi-sub-parameterized values
HSS (Smarandache, 2018) Able to deal with multi-sub-parametrization Unable to address uncertainty by multi-sub-parameterization
FHSS (Smarandache, 2018) Addresses FS with multi-sub-parameterized values Unable to address uncertainty exceeding FS’s restriction by multi-sub-parameterized values
IFHSS (Smarandache, 2018) Addresses IFS with multi-sub-parameterized values Unable to address uncertainty exceeding IFS’s restriction by multi-sub-parameterized values
q-ROFHSS (Khan, Gulistan & Wahab, 2022) Addresses q-ROFS with multi-sub-parameterized values Unable to address uncertainty exceeding q-ROFS’s restriction by multi-sub-parameterized values
q-RLDFHSS (Surya et al., 2024) Addresses q-RLDFS with multi-sub-parameterized values Unable to address uncertainty exceeding q-RLDFS’s restriction by multi-sub-parameterized values
LOq-RLDFHSS (proposed) Addresses q-RLDFHSS effectively even when there is a ranking among multi-sub-parameters Unable to address uncertainty exceeding q-RLDFS’s restriction by multi-sub-parameterized values
DOI: 10.7717/peerj-cs.2927/table-4

Discussion

Superiority of the proposed MADM method

In Table 4, the comparison analysis brings to light the exceptional superiority of the innovatively proposed MADM method when juxtaposed with the array of existing MADM methodologies rooted in fuzzy theories such as FS, IFS, PFS, q-ROFS, LDFS, q-RLDFS, SS, FSS, IFSS, q-ROFSS, LDFSS, HSS, FHSS, IFHSS, q-ROFHSS, LOSS, LOFSS and LOIFSS. The distinguished LOq-RLDFHSS based DM method, in its unmatched prowess, showcases its ability to effectively manage q-RLDFS even within the intricate complexities of multi-sub-attributed scenarios that entail the prioritization and ordering of these multi-sub-attributes. This distinctive characteristic proves to be more suitable in navigating through a wide spectrum of real-world MADM situations with finesse.

Computational effiency and scalability

The proposed MADM is highly efficient in terms of scalability since, it is capable of handling real-world problems with large data. Further, the proposed MADM methodology is capable of handling problems with large number of alternatives and parameters, but to understand the methodology clearly, the disaster management problem given in Section “MADM Approach Based on LOq-RLDFHSS” considers three alternatives and four parameters. Also, it is suitable to implement the proposed MADM method in various large-scale real-world applications such as medical diagnosis, supply chain optimization and more. In this study it is contributed to the field of disaster management. Also, the results obtained by the proposed method is more reliable and accurate since it considers more parameters and data, to handle the problem than the existing fuzzy MADM methods.

Contribution in the disaster management field

Even though various decision-making approaches and case studies (Arifah, Tariq & Juni, 2019; Wan et al., 2020; Mohammadnazari et al., 2022; Papathoma-Köhle et al., 2019) contribute to disaster management, those studies became inadequate when the disaster situation needed to incorporate more attributes together simultaneously to obtain the most appropriate solution. Further, the presented case study is a unique case in disaster management, which is not yet and unable to be described by the existing MADM approaches in the disaster management field. From this it becomes clear that conventional MADM strategies are inadequate when confronted with scenarios teeming with many intricate data, unlike our proposed method, which adeptly converts intricate parameter data into streamlined numerical formats.

Also, it is crucial to recognize that while the proposed method undeniably offers substantial benefits, it is not devoid of its own set of limitations, such as limitations mentioned in Table 4. Further, the algorithm shows a limitation in the case of ties.

Conclusion

For addressing a wide range of uncertain challenges, the q-RLDFHSS and LOq-RLDFHSS stand out as innovative extensions of FS theory. Throughout this research, numerous fundamental algebraic operations of LOq-RLDFHSS have been identified, emphasizing the development of an algorithm specifically designed to solve MADM problems leveraging the concepts of LOq-RLDFHSS. By exploring a unique MADM scenario within the domain of disaster management, which helps to choose the most appropriate plan to tackle the known upcoming natural disaster by considering more attributes together, the use of the suggested method in practice is thoroughly examined. The comparative analysis showcases the superiority and effectiveness of the novel MADM method against existing approaches, underscoring its value in real-world applications. In the comparative analysis, the study’s contribution to the disaster management field is also discussed in detail.

Future direction

In future, it will focus on developing advanced information measures and aggregation operators tailored for the LOq-RLDFHSS. Further, it will be focused on overcoming the limitations of the proposed study by utilizing the concept of hesitancy function described in Zia et al. (2024). Also, it is aimed to discuss various real-world problems in different domains such as medical, cybersecurity and pattern recognition.

Appendix

List of abbreviation used in the study

The list of most of the abbreviations used in this study is described in Table 5.

Table 5:
List of abbreviation used in the study.
Abbreviation Description
FS Fuzzy set
MADM Multi-attributed decision-making
MD Membership degree
IFS Intuitionistic fuzzy set
NMG Non-membership Degree
PFS Pythagorean fuzzy set
q-ROFS q-Rung orthopair fuzzy set
LDFS Linear Diophantine fuzzy set
RPs Reference parameters
q-RLDFS q-Rung linear Diophantine fuzzy set
SS Soft set
FSS Fuzzy soft set
IFSS Intuitonistic fuzzy soft set
q-ROFSS q-Rung orthopair fuzzy soft set
LDFSS Linear Diophantine fuzzy soft set
HSS Hypersoft set
FHSS Fuzzy hypersoft set
IFHSS Intuitionistic fuzzy hypersoft set
q-ROFHSS q-Rung orthopair fuzzy hypersoft set
q-RLDFHSS q-Rung linear diophantine fuzzy hypersoft set
LOSS Lattice ordered soft set
LOFSS Lattice ordered fuzzy soft set
LOIFSS Lattice ordered intuitionistic fuzzy soft set
LOq-RLDFSS Lattice ordered q-rung linear Diophantine fuzzy hypersoft set
DOI: 10.7717/peerj-cs.2927/table-5

Proof of proposition 3.2

Proof. Let (Θ1,1),(Θ2,2) LOq-RLDFHSS ( G). Then by Definition 3.2

Θ1(η)Θ2(η)=Θ3(η), where η3=12.

If 12=, then result is trivial.

Now for 12, since 1,2E1×E2×...×En

Therefore for any ηc1 ηd we have Θ1(ηc)Θ1(ηd),ηc,ηd1

and for any ςc2 ςd we have Θ2(ςc)Θ2(ςd),ςc,ςd2

Now for any ϖc,ϖd3 and ϖc3ϖd

ϖc,ϖd12

ϖc,ϖd1 and ϖc,ϖd2

Θ1(ϖc)Θ1(ϖd) and Θ2(ϖc)Θ2(ϖd) whenever ϖc1ϖd,ϖc2ϖd

ΩΘ1(ϖc)(g)ΩΘ1(ϖd)(g),ΩΘ2(ϖc)(g)ΩΘ2(ϖd)(g)

Θ1(ϖd)(g)Θ1(ϖc)(g),Θ2(ϖd)(g)Θ2(ϖc)(g)

ΔΘ1(ϖc)(g)ΔΘ1(ϖd)(g),ΔΘ2(ϖc)(g)ΔΘ2(ϖd)(g)

Θ1(ϖd)(g)Θ1(ϖc)(g),Θ2(ϖd)(g)Θ2(ϖc)(g)

Max {ΩΘ1(ϖc)(g),ΩΘ2(ϖc)(g)} Max {ΩΘ1(ϖd)(g),ΩΘ2(ϖd)(g)}

Min {Θ1(ϖd)(g),Θ2(ϖd)(g)} Min {Θ1(ϖc)(g),Θ2(ϖc)(g)}

Max {ΔΘ1(ϖc)(g),ΔΘ2(ϖc)(g)} Max {ΔΘ1(ϖd)(g),ΔΘ2(ϖd)(g)}

Min {Θ1(ϖd)(g),Θ2(ϖd)(g)} Min {Θ1(ϖc)(g),Θ2(ϖc)(g)}

ΩΘ1(ϖc)Θ2(ϖc)(g)ΩΘ1(ϖd)Θ2(ϖd)(g)

Θ1(ϖd)Θ2(ϖd)(g)Θ1(ϖc)Θ2(ϖc)(g)

ΔΘ1(ϖc)Θ2(ϖc)(g)ΔΘ1(ϖd)Θ2(ϖd)(g)

Θ1(ϖd)Θ2(ϖd)(g)Θ1(ϖc)Θ2(ϖc)(g)

ΩΘ3(ϖc)(g)ΩΘ3(ϖd)(g)

Θ3(ϖd)(g)Θ3(ϖc)(g)

ΔΘ3(ϖc)(g)ΔΘ3(ϖd)(g)

Θ3(ϖd)(g)Θ3(ϖc)(g)

Θ3(ϖc)Θ3(ϖd) for ϖc3ϖd

(Θ1,1)RES(Θ2,2) LOq-RLDFHSS (G). □

Proof of proposition 3.4

Proof. Let (Θ1,1),(Θ2,2) LOq-RLDFHSS ( G). Then by Definition 3.4

Θ1(η)Θ2(η)=Θ3(η), where η3=12.

If 12=, then result is trivial.

Now for 12, since 1,2E1×E2×...×En

Therefore for any ηc1 ηd we have Θ1(ηc)Θ1(ηd),ηc,ηd1

and for any ςc2ςd we have Θ2(ςc)Θ2(ςd),ςc,ςd2

Now for any ϖc,ϖd3 and ϖc3ϖd

ϖc,ϖd12

ϖc,ϖd1 and ϖc,ϖd2

Θ1(ϖc)Θ1(ϖd) and Θ2(ϖc)Θ2(ϖd) whenever ϖc1 ϖd,ϖc2ϖd

ΩΘ1(ϖc)(g)ΩΘ1(ϖd)(g),ΩΘ2(ϖc)(g)ΩΘ2(ϖd)(g)

Θ1(ϖd)(g)Θ1(ϖc)(g),Θ2(ϖd)(g)Θ2(ϖc)(g)

ΔΘ1(ϖc)(g)ΔΘ1(ϖd)(g),ΔΘ2(ϖc)(g)ΔΘ2(ϖd)(g)

Θ1(ϖd)(g)Θ1(ϖc)(g),Θ2(ϖd)(g)Θ2(ϖc)(g)

Min {ΩΘ1(ϖc)(g),ΩΘ2(ϖc)(g)} Min {ΩΘ1(ϖd)(g),ΩΘ2(ϖd)(g)}

Max {Θ1(ϖd)(g),Θ2(ϖd)(g)} Max {Θ1(ϖc)(g),Θ2(ϖc)(g)}

Min {ΔΘ1(ϖc)(g),ΔΘ2(ϖc)(g)} Min {ΔΘ1(ϖd)(g),ΔΘ2(ϖd)(g)}

Max {Θ1(ϖd)(g),Θ2(ϖd)(g)} Max {Θ1(ϖc)(g),Θ2(ϖc)(g)}

ΩΘ1(ϖc)Θ2(ϖc)(g)ΩΘ1(ϖd)Θ2(ϖd)(g)

Θ1(ϖd)Θ2(ϖd)(g)Θ1(ϖc)Θ2(ϖc)(g)

ΔΘ1(ϖc)Θ2(ϖc)(g)ΔΘ1(ϖd)Θ2(ϖd)(g)

Θ1(ϖd)Θ2(ϖd)(g)Θ1(ϖc)Θ2(ϖc)(g)

ΩΘ3(ϖc)(g)ΩΘ3(ϖd)(g)

Θ3(ϖd)(g)Θ3(ϖc)(g)

ΔΘ3(ϖc)(g)ΔΘ3(ϖd)(g)

Θ3(ϖd)(g)Θ3(ϖc)(g)

Θ3(ϖc)Θ3(ϖd) for ϖc 3 ϖd

(Θ1,1) RES (Θ2,2) LOq-RLDFHSS (G). □

Proof of proposition 3.6

Proof. Let (Θ1,1),(Θ2,2)LOq-RLDFHSS (G). Then by Definition 3.6

(Θ1,1)EXT(Θ2,2)=(Θ3,3) where 3=12

(Θ3,3)={{ΩΘ1(η)(g),Θ1(η)(g),ΔΘ1(η)(g),Θ1(η)(g)}ifη12{ΩΘ2(η)(g),Θ2(η)(g),ΔΘ2(η)(g),Θ2(η)(g)}ifη21{Max{ΩΘ1(η)(g),ΩΘ2(η)(g)},Min{Θ1(η)(g),Θ2(η)(g)},ifη12Max{ΔΘ1(η)(g),ΔΘ2(η)(g)},Min{Θ1(η)(g),Θ2(η)(g)}}suppose,

(Θ1,1)(Θ2,2). Then 12 and ΩΘ1(η)(g)ΩΘ2(η)(g),Θ2(η)(g)Θ1(η)(g),

ΔΘ1(η)(g)ΔΘ2(η)(g),Θ2(η)(g)Θ1(η)(g), for every η1 and gG

since 1,2E1×E2×...×En

Therefore for any ηc1 ηd we have Θ1(ηc)Θ1(ηd),ηc,ηd1

and for any ςc2ςd we have Θ2(ςc)Θ2(ςd),ςc,ςd2

Now for any ϖc,ϖd3 and ϖc3ϖd

ϖc,ϖd12

ϖc,ϖd12 or ϖc,ϖd2 and ϖc,ϖd1 because 12

now take ϖc,ϖd12

ϖc,ϖd1 and ϖc,ϖd2

Θ1(ϖc)Θ1(ϖd) and Θ2(ϖc)Θ2(ϖd) whenever ϖc1 ϖd,ϖc2ϖd

ΩΘ1(ϖc)(g)ΩΘ1(ϖd)(g),ΩΘ2(ϖc)(g)ΩΘ2(ϖd)(g)

Θ1(ϖd)(g)Θ1(ϖc)(g),Θ2(ϖd)(g)Θ2(ϖc)(g)

ΔΘ1(ϖc)(g)ΔΘ1(ϖd)(g),ΔΘ2(ϖc)(g)ΔΘ2(ϖd)(g)

Θ1(ϖd)(g)Θ1(ϖc)(g),Θ2(ϖd)(g)Θ2(ϖc)(g)

Max {ΩΘ1(ϖc)(g),ΩΘ2(ϖc)(g)} Max {ΩΘ1(ϖd)(g),ΩΘ2(ϖd)(g)}

Min {Θ1(ϖd)(g),Θ2(ϖd)(g)} Min {Θ1(ϖc)(g),Θ2(ϖc)(g)}

Max {ΔΘ1(ϖc)(g),ΔΘ2(ϖc)(g)} Max {ΔΘ1(ϖd)(g),ΔΘ2(ϖd)(g)}

Min {Θ1(ϖd)(g),Θ2(ϖd)(g)} Min {Θ1(ϖc)(g),Θ2(ϖc)(g)}

ΩΘ1(ϖc)Θ2(ϖc)(g)ΩΘ1(ϖd)Θ2(ϖd)(g)

Θ1(ϖd)Θ2(ϖd)(g)Θ1(ϖc)Θ2(ϖc)(g)

ΔΘ1(ϖc)Θ2(ϖc)(g)ΔΘ1(ϖd)Θ2(ϖd)(g)

Θ1(ϖd)Θ2(ϖd)(g)Θ1(ϖc)Θ2(ϖc)(g)

ΩΘ3(ϖc)(g)ΩΘ3(ϖd)(g)

Θ3(ϖd)(g)Θ3(ϖc)(g)

ΔΘ3(ϖc)(g)ΔΘ3(ϖd)(g)

Θ3(ϖd)(g)Θ3(ϖc)(g)

Θ3(ϖc)Θ3(ϖd) for ϖc3ϖd

Thus (Θ1,1)EXT(Θ2,2) LOq-RLDFHSS (G) if ϖc,ϖd12

Now suppose for any ϖc,ϖd2,ϖc,ϖd1 and ϖc2ϖd

Θ2(ϖc)Θ2(ϖd) whenever ϖc2ϖd

(Θ1,1)EXT(Θ2,2) LOq-RLDFHSS (G)

Hence (Θ1,1)EXT(Θ2,2) LOq-RLDFHSS (G) in both cases

(Θ1,1)EXT(Θ2,2) LOq-RLDFHSS (G), if one of them is a LOq-RLDFHS subset of other. □

Proof of proposition 3.9

Proof. Let (Θ1,1),(Θ2,2) LOq-RLDFHSS (G). Then by Definition 3.9

(Θ1,1)(Θ2,2)=(Ξ,1×2) also

Ξ(η,ς)(g)={Min{ΩΘ1(η)(g),ΩΘ2(ς)(g)},Max{Θ1(η)(g),Θ2(ς)(g)},

Min{ΔΘ1(η)(g),ΔΘ2(ς)(g)},Max{Θ1(η)(g),Θ2(ς)(g)}}

For any ηc1ηd we have Θ1(ηc)Θ1(ηd),ηc,ηd1

and for any ςc2ςd we have Θ2(ςc)Θ2(ςd),ςc,ςd2

Now for any (ηc,ςc),(ηd,ςd)1×2. Then by Definition 3.8

The order on 1×2 is (ηc,ςc)1 × 2 (ηd,ςd)ηc1ηd and ςc1ςd

Θ1(ηc)Θ1(ηd) and Θ2(ςc)Θ2(ςd)

ΩΘ1(ηc)(g)ΩΘ1(ηd)(g),ΩΘ2(ςc)(g)ΩΘ2(ςd)(g)

Θ1(ηd)(g)Θ1(ηc)(g),Θ2(ςd)(g)Θ2(ςc)(g)

ΔΘ1(ηc)(g)ΔΘ1(ηd)(g),ΔΘ2(ςc)(g)ΔΘ2(ςd)(g)

Θ1(ηd)(g)Θ1(ηc)(g),Θ2(ςd)(g)Θ2(ςc)(g)

Min {ΩΘ1(ηc)(g),ΩΘ2(ςc)(g)} Min {ΩΘ1(ηd)(g),ΩΘ2(ςd)(g)}

Max {Θ1(ηd)(g),Θ2(ςd)(g)} Max {Θ1(ηc)(g),Θ2(ςc)(g)}

Min {ΔΘ1(ηc)(g),ΔΘ2(ςc)(g)} Min {ΔΘ1(ηd)(g),ΔΘ2(ςd)(g)}

Max {Θ1(ηd)(g),Θ2(ςd)(g)} Max {Θ1(ηc)(g),Θ2(ςc)(g)}

ΩΞ(ηc,ςc)(g)ΩΞ(ηd,ςd)(g)

Ξ(ηd,ςd)(g)Ξ(ηc,ςc)(g)

ΔΞ(ηc,ςc)(g)ΔΞ(ηd,ςd)(g)

Ξ(ηd,ςd)(g)Ξ(ηc,ςc)(g)

Ξ(ηc,ςc)Ξ(ηd,ςd) for (ηc,ςc)1×2(ηd,ςd)

Therefore, (Θ1,1)(Θ2,2) LOq-RLDFHSS (G). □

Proof of proposition 3.11

Proof. Let (Θ1,1),(Θ2,2) LOq-RLDFHSS (G). Then by Definition 3.11

(Θ1,1)(Θ2,2)=(Ξ,1×2) also

Ξ(η,ς)(g)={Max{ΩΘ1(η)(g),ΩΘ2(ς)(g)},Min{Θ1(η)(g),Θ2(ς)(g)},

Max{ΔΘ1(η)(g),ΔΘ2(ς)(g)},Min{Θ1(η)(g),Θ2(ς)(g)}}

For any ηc1ηd we have Θ1(ηc)Θ1(ηd),ηc,ηd1

and for any ςc2ςd we have Θ2(ςc)Θ2(ςd),ςc,ςd2

Now for any (ηc,ςc),(ηd,ςd)1×2. Then by Definition 3.8

Now for any (ηc,ςc),(ηd,ςd)1×2. Then by Definition 3.8

The order on 1×2 is (ηc,ςc)1 × 2 (ηd,ςd)ηc1ηd and ςc1ςd

Θ1(ηc)Θ1(ηd) and Θ2(ςc)Θ2(ςd)

ΩΘ1(ηc)(g)ΩΘ1(ηd)(g),ΩΘ2(ςc)(g)ΩΘ2(ςd)(g)

Θ1(ηd)(g)Θ1(ηc)(g),Θ2(ςd)(g)Θ2(ςc)(g)

ΔΘ1(ηc)(g)ΔΘ1(ηd)(g),ΔΘ2(ςc)(g)ΔΘ2(ςd)(g)

Θ1(ηd)(g)Θ1(ηc)(g),Θ2(ςd)(g)Θ2(ςc)(g)

Max {ΩΘ1(ηc)(g),ΩΘ2(ςc)(g)} Max {ΩΘ1(ηd)(g),ΩΘ2(ςd)(g)}

Min {Θ1(ηd)(g),Θ2(ςd)(g)} Min {Θ1(ηc)(g),Θ2(ςc)(g)}

Max {ΔΘ1(ηc)(g),ΔΘ2(ςc)(g)} Max {ΔΘ1(ηd)(g),ΔΘ2(ςd)(g)}

Min {Θ1(ηd)(g),Θ2(ςd)(g)} Min {Θ1(ηc)(g),Θ2(ςc)(g)}

ΩΞ(ηc,ςc)(g)ΩΞ(ηd,ςd)(g)

Ξ(ηd,ςd)(g)Ξ(ηc,ςc)(g)

ΔΞ(ηc,ςc)(g)ΔΞ(ηd,ςd)(g)

Ξ(ηd,ςd)(g)Ξ(ηc,ςc)(g)

Ξ(ηc,ςc)Ξ(ηd,ςd) for (ηc,ςc)1×2(ηd,ςd)

Therefore, (Θ1,1)(Θ2,2) LOq-RLDFHSS (G). □