Volume 2012, Issue 1 938910
Research Article
Open Access

Generalizations of 𝒩-Subalgebras in BCK/BCI-Algebras Based on Point 𝒩-Structures

Young Bae Jun

Young Bae Jun

Department of Mathematics Education (and RINS), Gyengsang National University, Jinju 660-701, Republic of Korea gnu.ac.kr

Search for more papers by this author
Kyoung Ja Lee

Kyoung Ja Lee

Department of Mathematics Education, Hannam University, Daejeon 306-791, Republic of Korea hannam.ac.kr

Search for more papers by this author
Min Su Kang

Corresponding Author

Min Su Kang

Department of Mathematics, Hanyang University, Seoul 133-791, Republic of Korea hanyang.ac.kr

Search for more papers by this author
First published: 04 October 2012
Academic Editor: Nicola Mastronardi

Abstract

The aim of this article is to obtain more general forms than the papers of (Jun et al. (2010); Jun et al. (in press)). The notions of 𝒩-subalgebras of types (∈, qk), (∈, ∈∨qk), and (q, ∈∨qk) are introduced, and the concepts of   qk-support and ∈∨qk-support are also introduced. Several related properties are investigated. Characterizations of 𝒩-subalgebra of type (∈, ∈∨qk) are discussed, and conditions for an 𝒩-subalgebra of type (∈, ∈∨qk) to be an 𝒩-subalgebra of type (∈, ∈) are considered.

1. Introduction

A (crisp) set A in a universe X can be defined in the form of its characteristic function ΞΌA : X β†’ {0,1} yielding the value 1 for elements belonging to the set A and the value 0 for elements excluded from the set A. So far most of the generalizations of the crisp set have been conducted on the unit interval [0,1], and they are consistent with the asymmetry observation. In other words, the generalization of the crisp set to fuzzy sets relied on spreading positive information that fits the crisp point {1} into the interval [0,1]. Because no negative meaning of information is suggested, we now feel a need to deal with negative information. To do so, we also feel a need to supply mathematical tool. To attain such object, Jun et al. [1] introduced a new function which is called negative-valued function and constructed 𝒩-structures. They applied 𝒩-structures to BCK/BCI-algebras and discussed 𝒩-subalgebras and 𝒩-ideals in BCK/BCI-algebras. Jun et al. [2] considered closed ideals in BCH-algebras based on 𝒩-structures. To obtain more general form of an 𝒩-subalgebra in BCK/BCI-algebras, Jun et al. [3] defined the notions of 𝒩-subalgebras of types (∈, ∈), (∈, q), (∈, ∈∨q), (q, ∈), (q, q), and (q, ∈∨q) and investigated related properties. They also gave conditions for an 𝒩-structure to be an 𝒩-subalgebra of type (q, ∈∨q). Jun et al. provided a characterization of an 𝒩-subalgebra of type (∈, ∈∨q) (see [3, 4]).

In this paper, we try to have more general form of the papers [3, 4]. We introduce the notions of 𝒩-subalgebras of types (∈, qk), (∈, ∈∨qk), and (q, ∈∨qk). We also introduce the concepts of qk-support and ∈∨qk-support and investigate several properties. We discuss characterizations of 𝒩-subalgebra of type (∈, ∈∨qk). We consider conditions for an 𝒩-subalgebra of type (∈, ∈∨qk) to be an 𝒩-subalgebra of type (∈, ∈). The important achievement of the study of 𝒩-subalgebras of types (∈, qk), (∈, ∈∨qk), and (q, ∈∨qk) is that the notions of 𝒩-subalgebras of types (∈, q), (∈, ∈∨q), and (q, ∈∨q) are a special case of 𝒩-subalgebras of types (∈, qk), (∈, ∈∨qk), and (q, ∈∨qk), and thus so many results in the papers [3, 4] are corollaries of our results obtained in this paper.

2. Preliminaries

Let K(Ο„) be the class of all algebras with type Ο„ = (2,0). By a BCI-algebra, we mean a system X : = (X, *, 0) ∈ K(Ο„) in which the following axioms hold:
  • (i)

    ((x*y)*(x*z))*(z*y) = 0,

  • (ii)

    (x*(x*y))*y = 0,

  • (iii)

    x*x = 0,

  • (iv)

    x*y = y*x =   0β‡’x = y,

for all x, y, z ∈ X. If a BCI-algebra X satisfies 0*x = 0 for all x ∈ X, then we say that X is a BCK-algebra. We can define a partial ordering ≀ by
()
In a BCK/BCI-algebra X, the following hold:
  •  

    (a1) (βˆ€x ∈ X)(x*0 = x),

  •  

    (a2) (βˆ€x, y, z ∈ X)((x*y)*z = (x*z)*y),

for all x, y, z ∈ X.

A nonempty subset S of a BCK/BCI-algebras X is called a subalgebra of X if x*y ∈ S for all x, y ∈ S. For our convenience, the empty set βˆ… is regarded as a subalgebra of X.

We refer the reader to the books [5, 6] for further information regarding BCK/BCI-algebras.

For any family {ai∣i ∈ Ξ›} of real numbers, we define
()

Denote by β„±(X, [βˆ’1,0]) the collection of functions from a set X to [βˆ’1,0]. We say that an element of β„±(X, [βˆ’1,0]) is a negative-valued function from X to [βˆ’1,0] (briefly, 𝒩-function on X). By an 𝒩-structure, we mean an ordered pair (X, f) of X and an 𝒩-function f on X. In what follows, let X denote a BCK/BCI-algebras and f an 𝒩-function on X unless otherwise specified.

Definition 2.1 (see [1].)By a subalgebra of X based on 𝒩-function f (briefly, 𝒩-subalgebra of X), we mean an 𝒩-structure (X, f) in which f satisfies the following assertion:

()

For any 𝒩-structure (X, f) and t ∈ [βˆ’1,0), the set

()
is called a closed  t-support of (X, f), and the set
()
is called an open   t-support of (X, f).

Using the similar method to the transfer principle in fuzzy theory (see [7, 8]), Jun et al. [2] considered transfer principle in 𝒩-structures as follows.

Theorem 2.2 (see [2]; 𝒩-transfer principle.)An 𝒩-structure (X, f) satisfies the property if and only if for all Ξ± ∈ [βˆ’1,0],

()

Lemma 2.3 (see [1].)An 𝒩-structure (X, f) is an 𝒩-subalgebra of X if and only if every open t-support of (X, f) is a subalgebra of X for all t ∈ [βˆ’1,0).

3. General Form of 𝒩-Subalgebras with Type (∈, ∈∨q)

In what follows, let t and k denote arbitrary elements of [βˆ’1,0) and (βˆ’1,0], respectively, unless otherwise specified.

Let (X, f) be an 𝒩-structure in which f is given by
()
In this case, f is denoted by xt, and we call (X, xt) a point 𝒩-structure. For any 𝒩-structure (X, g), we say that a point 𝒩-structure (X, xt) is an π’©βˆˆ-subset (resp., 𝒩q-subset) of (X, g) if g(x) ≀ t (resp., g(x) + t + 1 < 0). If a point 𝒩-structure (X, xt) is an π’©βˆˆ-subset of (X, g) or an 𝒩q-subset of (X, g), we say (X, xt) is an π’©βˆˆβˆ¨q-subset of (X, g). We say that a point 𝒩-structure (X, xt) is an -subset of (X, g) if g(x) + t βˆ’ k + 1 < 0. Clearly, every -subset with k = 0 is an 𝒩q-subset. Note that if k, r ∈ (βˆ’1,0] with k < r, then every -subset is an -subset.

Definition 3.1. An 𝒩-structure (X, f) is called an 𝒩-subalgebra of type

  • (i)

    (∈, ∈) (resp., (∈, q) and (∈, ∈∨q)) if whenever two point 𝒩-structures and are π’©βˆˆ-subsets of (X, f) then the point 𝒩-structure is an π’©βˆˆ-subset (resp., 𝒩q-subset and π’©βˆˆβˆ¨q-subset) of (X, f).

  • (ii)

    (q, ∈) (resp., (q, q) and (q, ∈∨q)) if whenever two point 𝒩-structures and are 𝒩q-subsets of (X, f) then the point 𝒩-structure is an π’©βˆˆ-subset (resp., 𝒩q-subset and π’©βˆˆβˆ¨q-subset) of (X, f).

Definition 3.2. An 𝒩-structure (X, f) is called an 𝒩-subalgebra of type (∈, ∈∨qk) (resp., ( q, ∈∨qk)) if whenever two point 𝒩-structures and are π’©βˆˆ-subsets (resp., 𝒩q-subsets) of (X, f) then the point 𝒩-structure is an -subset of (X, f).

Example 3.3. Consider a BCI-algebra X = {0, a, b, c} with the following Cayley table:

()
Let (X, f) be an 𝒩-structure in which f is defined by
()
It is routine to verify that (X, f) is an 𝒩-subalgebra of type (∈, ∈∨qβˆ’0.2).

Note that if k, r ∈ (βˆ’1,0] with k < r, then every 𝒩-subalgebra of type (∈, ∈∨qk) is an 𝒩-subalgebra of type (∈, ∈∨qr), but the converse is not true as seen in the following example.

Example 3.4. The 𝒩-subalgebra (X, f) of type (∈, ∈∨qβˆ’0.2) in Example 3.3 is not of type (∈, ∈∨qβˆ’0.4) since (X, aβˆ’0.65) and (X, aβˆ’0.68) are π’©βˆˆ-subsets of (X, f), but

()
is not an -subset of (X, f).

Theorem 3.5. Every 𝒩-subalgebra of type (∈, ∈) is of type (∈, ∈∨qk).

Proof. Straightforward.

Taking k = 0 in Theorem 3.5 induces the following corollary.

Corollary 3.6. Every 𝒩-subalgebra of type (∈, ∈) is of type (∈, ∈∨q).

The converse of Theorem 3.5 is not true as seen in the following example.

Example 3.7. Consider the 𝒩-subalgebra (X, f) of type (∈, ∈∨qβˆ’0.2) which is given in Example 3.3. Then (X, f) is not an 𝒩-subalgebra of type (∈, ∈) since (X, aβˆ’0.65) and (X, aβˆ’0.68) are π’©βˆˆ-subsets of (X, f), but (X, (a*a) ⋁{βˆ’0.65,βˆ’0.68}) is not an π’©βˆˆ-subset of (X, f).

Definition 3.8. An 𝒩-structure (X, f) is called an 𝒩-subalgebra of type (∈, qk) if whenever two point 𝒩-structure and are π’©βˆˆ-subsets of (X, f) then the point 𝒩-structure is an -subset of (X, f).

Theorem 3.9. Every 𝒩-subalgebra of type (∈, qk) is of type (∈, ∈∨qk).

Proof. Straightforward.

Taking k = 0 in Theorem 3.9 induces the following corollary.

Corollary 3.10. Every 𝒩-subalgebra of type (∈, q) is of type (∈, ∈∨q).

The converse of Theorem 3.9 is not true as seen in the following example.

Example 3.11. Consider the 𝒩-subalgebra (X, f) of type (∈, ∈∨qβˆ’0.2) which is given in Example 3.3. Then (X, aβˆ’0.65) and (X, bβˆ’0.25) are 𝒩-subsets of (X, f), but

()
is not an -subset of (X, f) for k = βˆ’0.2 since f(c) βˆ’ 0.25 βˆ’ 0.2 + 1 > 0.

We consider a characterization of an 𝒩-subalgebra of type (∈, ∈∨qk).

Theorem 3.12. An 𝒩-structure (X, f) is an 𝒩-subalgebra of type (∈, ∈∨qk) if and only if it satisfies

()

Proof. Let (X, f) be an 𝒩-structure of type (∈, ∈∨qk). Assume that (3.6) is not valid. Then there exists a, b ∈ X such that

()
If ⋁{f(a), f(b)} > (k βˆ’ 1)/2, then f(a*b) > ⋁{f(a), f(b)}. Hence
()
for some t ∈ [βˆ’1,0). It follows that point 𝒩-structures (X, at) and (X, bt) are π’©βˆˆ-subsets of (X, f), but the point 𝒩-structure (X, (a*b) t) is not an π’©βˆˆ-subset of (X, f). Moreover,
()
and so (X, (a*b) t) is not an -subset of (X, f). Consequently, (X, (a*b) t) is not an -subset of (X, f). This is a contradiction. If ⋁{f(a), f(b)} ≀ (k βˆ’ 1)/2, then f(a) ≀ (k βˆ’ 1)/2, f(b) ≀ (k βˆ’ 1)/2 and f(a*b) > (k βˆ’ 1)/2. Thus (X, a(k βˆ’ 1)/2) and (X, b(k βˆ’ 1)/2) are π’©βˆˆ-subsets of (X, f), but (X, (a*b) (k βˆ’ 1)/2) is not an π’©βˆˆ-subset of (X, f). Also,
()
that is, (X, (a*b) (k βˆ’ 1)/2) is not an -subset of (X, f). Hence (X, (a*b) (k βˆ’ 1)/2) is not an -subset of (X, f), a contradiction. Therefore (3.6) is valid.

Conversely, suppose that (3.6) is valid. Let x, y ∈ X and t1, t2 ∈ [βˆ’1,0) be such that two point 𝒩-structures and are π’©βˆˆ-subsets of (X, f). Then

()
Assume that t1 β‰₯ (k βˆ’ 1)/2 or t2 β‰₯ (k βˆ’ 1)/2. Then f(x*y) ≀ ⋁{t1, t2}, and so is an π’©βˆˆ-subset of (X, f). Now suppose that t1 < (k βˆ’ 1)/2 and t2 < (k βˆ’ 1)/2. Then f(x*y) ≀ (k βˆ’ 1)/2, and thus
()
that is, is an -subset of (X, f). Therefore is an -subset of (X, f) and consequently (X, f) is an 𝒩-subalgebra of type (∈, ∈∨qk).

Corollary 3.13 (see [3].)An 𝒩-structure (X, f) is an 𝒩-subalgebra of type (∈, ∈∨q) if and only if it satisfies

()

Proof. It follows from taking k = 0 in Theorem 3.12.

We provide conditions for an 𝒩-structure to be an 𝒩-subalgebra of type (q, ∈∨qk).

Theorem 3.14. Let S be a subalgebra of X and let (X, f) be an 𝒩-structure such that

  • (a)

    (βˆ€x ∈ X)(x ∈ Sβ‡’f(x)≀(k βˆ’ 1)/2),

  • (b)

    (βˆ€x ∈ X)(x βˆ‰ Sβ‡’f(x) = 0).

Then (X, f) is an 𝒩-subalgebra of type (q, ∈∨qk).

Proof. Let x, y ∈ X and t1, t2 ∈ [βˆ’1,0) be such that two point 𝒩-structures and are 𝒩q-subsets of (X, f). Then f(x) + t1 + 1 < 0 and f(y) + t2 + 1 < 0. Thus x*y ∈ S because if it is impossible, then x βˆ‰ S or y βˆ‰ S. Thus f(x) = 0 or f(y) = 0, and so t1 < βˆ’1 or t2 < βˆ’1. This is a contradiction. Hence f(x*y)≀(k βˆ’ 1)/2. If ⋁{t1, t2} < (k βˆ’ 1)/2, then f(x*y) + ⋁{t1, t2} βˆ’ k + 1 < ((k βˆ’ 1)/2) + ((k βˆ’ 1)/2) βˆ’ k + 1 = 0 and so the point 𝒩-structure is an -subset of (X, f). If ⋁{t1, t2} β‰₯ (k βˆ’ 1)/2, then f(x*y)≀(k βˆ’ 1)/2 ≀ ⋁{t1, t2} and so the point 𝒩-structure is an π’©βˆˆ-subset of (X, f). Therefore the point 𝒩-structure is an -subset of (X, f). This shows that (X, f) is an 𝒩-subalgebra of type (q, ∈∨qk).

Taking k = 0 in Theorem 3.14, we have the following corollary.

Corollary 3.15 (see [3].)Let S be a subalgebra of X and let (X, f) be an 𝒩-structure such that

  • (a)

    (βˆ€x ∈ X)(x ∈ Sβ‡’f(x) ≀ βˆ’0.5),

  • (b)

    (βˆ€x ∈ X)(x βˆ‰ Sβ‡’f(x) = 0).

Then (X, f) is an 𝒩-subalgebra of type (q, ∈∨q).

Theorem 3.16. Let (X, f) be an 𝒩-subalgebra of type (qk, ∈∨qk). If f is not constant on the open 0-support of (X, f), then f(x) ≀ (k βˆ’ 1)/2 for some x ∈ X. In particular, f(0) ≀ (k βˆ’ 1)/2.

Proof. Assume that f(x)>(k βˆ’ 1)/2 for all x ∈ X. Since f is not constant on the open 0-support of (X, f), there exists x ∈ O(f; 0) such that tx = f(x) β‰  f(0) = t0. Then either t0 < tx or t0 > tx. For the case t0 < tx, choose r < (k βˆ’ 1)/2 such that t0 + r βˆ’ k + 1 < 0 < tx + r βˆ’ k + 1. Then the point 𝒩-structure (X, 0r) is an -subset of (X, f). Since (X, xβˆ’1) is an -subset of (X, f). It follows from (a1) that the point 𝒩-structure (X, (x*0) ⋁{r,βˆ’1}) = (X, xr) is an -subset of (X, f). But, f(x) > (k βˆ’ 1)/2 > r implies that the point 𝒩-structure (X, xr) is not an π’©βˆˆ-subset of (X, f). Also, f(x) + r βˆ’ k + 1 = tx + r βˆ’ k + 1 > 0 implies that the point 𝒩-structure (X, xr) is not an -subset of (X, f). This is a contradiction. Assume that t0 > tx and take r < (k βˆ’ 1)/2 such that tx + r βˆ’ k + 1 < 0 < t0 + r βˆ’ k + 1. Then (X, xr) is an -subset of (X, f). Since

()
(X, (x*x) ⋁{r,r}) is not an π’©βˆˆ-subset of (X, f). Since
()
(X, (x*x) ⋁{r,r}) is not an -subset of (X, f). Hence (X, (x*x) ⋁{r,r}) is not an -subset of (X, f), which is a contradiction. Therefore f(x)≀(k βˆ’ 1)/2 for some x ∈ X. We now prove that f(0) ≀ (k βˆ’ 1)/2. Assume that f(0) = t0 > (k βˆ’ 1)/2. Note that there exists x ∈ X such that f(x) = tx ≀ (k βˆ’ 1)/2 and so tx < t0. Choose t1 < t0 such that tx + t1 βˆ’ k + 1 < 0 < t0 + t1 βˆ’ k + 1. Then f(x) + t1 βˆ’ k + 1 = tx + t1 βˆ’ k + 1 < 0, and thus the point 𝒩-structure is an -subset of (X, f). Now we have
()
and f(x*x) = f(0) = t0 > t1 = ⋁{t1, t1}. Hence is not an -subset of (X, f). This is a contradiction, and therefore f(0)≀(k βˆ’ 1)/2.

Corollary 3.17 (see [3].)Let (X, f) be an 𝒩-subalgebra of type (q, ∈∨q). If f is not constant on the open 0-support of (X, f), then f(x)β‰€βˆ’0.5 for some x ∈ X. In particular, f(0)β‰€βˆ’0.5.

Theorem 3.18. An 𝒩-structure (X, f) is an 𝒩-subalgebra of type (∈, ∈∨qk) if and only if for every t ∈ [(k βˆ’ 1)/2, 0] the nonempty closed t-support of (X, f) is a subalgebra of X.

Proof. Assume that (X, f) is an 𝒩-subalgebra of type (∈, ∈∨qk) and let t ∈ [(k βˆ’ 1)/2, 0] be such that C(f; t) β‰  βˆ…. Let x, y ∈ C(f; t). Then f(x) ≀ t and f(y) ≀ t. It follows from Theorem 3.12 that

()
so that x*y ∈ C(f; t). Therefore C(f; t) is a subalgebra of X.

Conversely, let (X, f) be an 𝒩-structure such that the nonempty closed t-support of (X, f) is a subalgebra of X for all t ∈ [(k βˆ’ 1)/2, 0]. If there exist a, b ∈ X such that f(a*b) > ⋁{f(a), f(b), (k βˆ’ 1)/2}, then we can take s ∈ [βˆ’1,0] such that

()
Thus a, b ∈ C(f; s) and s β‰₯ (k βˆ’ 1)/2. Since C(f, s) is a subalgebra of X, it follows that a*b ∈ C(f; s) so that f(a*b) ≀ s. This is a contradiction, and therefore f(x*y) ≀ ⋁{f(x), f(y), (k βˆ’ 1)/2} for all x, y ∈ X. Using Theorem 3.12, we conclude that (X, f) is an 𝒩-subalgebra of type (∈, ∈∨qk).

Taking k = 0 in Theorem 3.18, we have the following corollary.

Corollary 3.19 (see [4].)An 𝒩-structure (X, f) is an 𝒩-subalgebra of type (∈, ∈∨q) if and only if for every t ∈ [βˆ’0.5,0] the nonempty closed t-support of (X, f) is a subalgebra of X.

Theorem 3.20. Let S be a subalgebra of X. For any t ∈ [(k βˆ’ 1)/2, 0), there exists an 𝒩-subalgebra (X, f) of type (∈, ∈∨qk) for which S is represented by the closed t-support of (X, f).

Proof. Let (X, f) be an 𝒩-structure in which f is given by

()
for all x ∈ X where t ∈ [(k βˆ’ 1)/2, 0). Assume that f(a*b) > ⋁{f(a), f(b), (k βˆ’ 1)/2} for some a, b ∈ X. Since the cardinality of the image of f is 2, we have f(a*b) = 0 and ⋁{f(a), f(b), (k βˆ’ 1)/2} = t. Since t β‰₯ (k βˆ’ 1)/2, it follows that f(a) = t = f(b) so that a, b ∈ S. Since S is a subalgebra of X, we obtain a*b ∈ S and so f(a*b) = t < 0. This is a contradiction. Therefore f(x*y) ≀ ⋁{f(x), f(y), (k βˆ’ 1)/2} for all x, y ∈ X. Using Theorem 3.12, we conclude that (X, f) is an 𝒩-subalgebra of type (∈, ∈∨qk). Obviously, S is represented by the closed t-support of (X, f).

Corollary 3.21 (see [4].)Let S be a subalgebra of X. For any t ∈ [βˆ’0.5,0), there exists an 𝒩-subalgebra (X, f) of type (∈, ∈∨q) for which S is represented by the closed t-support of (X, f).

Proof. It follows from taking k = 0 in Theorem 3.20.

Note that every 𝒩-subalgebra of type (∈, ∈) is an 𝒩-subalgebra of type (∈, ∈∨qk), but the converse is not true in general (see Example 3.7). Now, we give a condition for an 𝒩-subalgebra of type (∈, ∈∨qk) to be an 𝒩-subalgebra of type (∈, ∈).

Theorem 3.22. Let (X, f) be an 𝒩-subalgebra of type (∈, ∈∨qk) such that f(x)>(k βˆ’ 1)/2  for all x ∈ X. Then (X, f) is an 𝒩-subalgebra of type (∈, ∈).

Proof. Let x, y ∈ X and t ∈ [βˆ’1,0) be such that and are π’©βˆˆ-subsets of (X, f). Then f(x) ≀ t1 and f(y) ≀ t2. It follows from Theorem 3.12 and the hypothesis that

()
so that is an π’©βˆˆ-subset of (X, f). Therefore (X, f) is an 𝒩-subalgebra of type (∈, ∈).

Corollary 3.23 (see [4].)Let (X, f) be an 𝒩-structure of type (∈, ∈∨q) such that f(x)>βˆ’0.5 for all x ∈ X. Then (X, f) is an 𝒩-subalgebra of type (∈, ∈).

Proof. It follows from taking k = 0 in Theorem 3.22.

Theorem 3.24. Let {(X, fi)∣i ∈ Ξ›} be a family of 𝒩-subalgebras of type (∈, ∈∨qk). Then (X, ⋃iβˆˆΞ›fi) is an 𝒩-subalgebra of type (∈, ∈∨qk), where ⋃iβˆˆΞ›fi is an 𝒩-function on X given by (⋃iβˆˆΞ›fi)(x) = ⋁iβˆˆΞ›fi(x) for all x ∈ X.

Proof. Let x, y ∈ X and t1, t2 ∈ [βˆ’1,0) be such that and are π’©βˆˆ-subsets of (X, ⋃iβˆˆΞ›fi). Assume that is not an -subset of (X, ⋃iβˆˆΞ›fi). Then is neither an π’©βˆˆ-subset nor an -subset of (X, ⋃iβˆˆΞ›fi). Hence (⋃iβˆˆΞ›fi)(x*y) > ⋁{t1, t2} and

()
which imply that
()
Let is an π’©βˆˆ-subset of  (X, fi)} and is an is not an π’©βˆˆ-subset  of  (X, fj)}. Then Ξ› = A1 βˆͺ A2 and A1∩A2 = βˆ…. If A2 = βˆ…, then is an π’©βˆˆ-subset of (X, fi) for all i ∈ Ξ›, that is, fi(x*y) ≀ ⋁{t1, t2} for all i ∈ Ξ›. Thus (⋃iβˆˆΞ›fi)(x*y) ≀ ⋁{t1, t2}. This is a contradiction. Hence A2 β‰  βˆ…, and so for every i ∈ A2, we have fi(x*y) > ⋁{t1, t2} and fi(x*y) + ⋁{t1, t2} βˆ’ k + 1 < 0. It follows that ⋁{t1, t2} < (k βˆ’ 1)/2. Since is an π’©βˆˆ-subset of (X, ⋃iβˆˆΞ›fi), we have
()
for all i ∈ Ξ›. Similarly, fi(y)<(k βˆ’ 1)/2 for all i ∈ Ξ›. Next suppose that t : = fi(x*y) > (k βˆ’ 1)/2. Taking (k βˆ’ 1)/2 < r < t, we know that (X, xr) and (X, yr) are π’©βˆˆ-subsets of (X, fi), but (X, (x*y) ⋁{r,r}) = (X, (x*y) r) is not an -subset of (X, fi). This contradicts that (X, fi) is an 𝒩-subalgebra of type (∈, ∈∨qk). Hence fi(x*y) ≀ (k βˆ’ 1)/2 for all i ∈ Ξ›, and so (⋃iβˆˆΞ›fi)(x*y)≀(k βˆ’ 1)/2 which contradicts (3.22). Therefore is an -subset of (X, ⋃iβˆˆΞ›fi) and consequently (X, ⋃iβˆˆΞ›fi) is an 𝒩-subalgebra of type (∈, ∈∨qk).

For any 𝒩-structure (X, f) and t ∈ [βˆ’1,0), the q-support and the ∈∨q-support of (X, f) related to t are defined to be the sets (see [4])
()
()
respectively. Note that the ∈∨q-support is the union of the closed support and the q-support, that is,
()
The qk-support and the ∈∨qk-support of (X, f) related to t are defined to be the sets
()
()
respectively. Clearly, for all t ∈ [βˆ’1,0).

Theorem 3.25. An 𝒩-structure (X, f) is an 𝒩-subalgebra of type (∈, ∈∨qk) if and only if the ∈∨qk-support of (X, f) related to t is a subalgebra of X for all t ∈ [βˆ’1,0).

Proof. Suppose that (X, f) is an 𝒩-subalgebra of type (∈, ∈∨qk). Let for t ∈ [βˆ’1,0). Then (X, xt) and (X, yt) are -subsets of (X, f). Hence f(x) ≀   t or f(x) +   t  βˆ’   k  +   1   <  0, and f(y) ≀ t or f(y) + t βˆ’ k + 1 < 0. Then we consider the following four cases:

  •  

    (c1) f(x) ≀ t and f(y) ≀ t,

  •  

    (c2) f(x) ≀ t and f(y) + t βˆ’ k + 1 < 0,

  •  

    (c3) f(x) + t βˆ’ k + 1 < 0 and f(y) ≀ t,

  •  

    (c4) f(x) + t βˆ’ k + 1 < 0 and f(y) + t βˆ’ k + 1 < 0.

Combining (3.6) and (c1), we have f(x*y) ≀ ⋁{t, (k βˆ’ 1)/2}. If t β‰₯ (k βˆ’ 1)/2, then f(x*y) ≀ t and so (X, (x*y) t) is an π’©βˆˆ-subset of (X, f). Hence . If t < (k βˆ’ 1)/2, then f(x *  y)≀(k βˆ’ 1)/2 and so f(x *  y) + t βˆ’ k + 1 < ((k βˆ’ 1)/2)+((k βˆ’ 1)/2) βˆ’ k + 1 = 0, that is, (X, (x*y) t) is an -subset of (X, f). Therefore . For the case (c2), assume that t < (k βˆ’ 1)/2. Then
()
and so f(x*y) + t βˆ’ k + 1 < 0. Thus (X, (x*y) t) is an -subset of (X, f). If t β‰₯ (k βˆ’ 1)/2, then
()
and thus or x*y ∈ C(f; t). Consequently, . For the case (c3), it is similar to the case (c2). Finally, for the case (c4), if t β‰₯ (k βˆ’ 1)/2, then k βˆ’ 1 βˆ’ t ≀ (k βˆ’ 1)/2 ≀ t. Hence
()
which implies that x*y ∈ C(f; t). If t < (k βˆ’ 1)/2, then t < (k βˆ’ 1)/2 < k βˆ’ 1 βˆ’ t. Therefore
()
that is, f(x*y) + t βˆ’ k + 1 < 0, which means that (X, (x*y) t) is an -subset of (X, f). Consequently, the ∈∨qk-support of (X, f) related to t is a subalgebra of X for all t ∈ [βˆ’1,0).

Conversely, let (X, f) be an 𝒩-structure for which the ∈∨qk-support of (X, f) related to t is a subalgebra of X for all t ∈ [βˆ’1,0). Assume that there exist a, b ∈ X such that f(a*b) > ⋁{f(a), f(b), (k βˆ’ 1)/2}. Then

()
for some s ∈ [(k βˆ’ 1)/2, 0). It follows that but a*b βˆ‰ C(f; s). Also, f(a*b) + s βˆ’ k + 1 > 2s βˆ’ k + 1 β‰₯ 0, that is, . Thus which is a contradiction. Therefore
()
for all x, y ∈ X. Using Theorem 3.12, we conclude that (X, f) is an 𝒩-subalgebra of type (∈, ∈∨qk).

If we take k = 0 in Theorem 3.25, we have the following corollary.

Corollary 3.26 (see [4].)An 𝒩-structure (X, f) is an 𝒩-subalgebra of type (∈, ∈∨q) if and only if the ∈∨q-support of (X, f) related to t is a subalgebra of X for all t ∈ [βˆ’1,0).

    The full text of this article hosted at iucr.org is unavailable due to technical difficulties.