On the algebraic structure of binary lattice-valued fuzzy relations

From a general algebraic point of view, this paper aims at providing an algebraic analysis for binary lattice-valued relations based on lattice implication algebras—a kind of lattice-valued propositional logical algebra. By abstracting away from the concrete lattice-valued relations and the operations on them, such as composition and converse, the notion of lattice-valued relation algebra is introduced, LRA for short. The reduct of an LRA is a lattice implication algebra. Such an algebra generalizes Boolean relation algebras by general distributive lattices and can provide a fundamental algebraic theory for establishing lattice-valued first-order logic. Some important results are generalized from the classical case. The notion of cylindric filter is introduced and the generated cylindric filters are characterized.     

Algebraic Properties of Cores of Generalized Neurofunctions

This paper considers generalized neural elements and identifies the conditions for implementation of functions of algebra of logic from these elements. We introduce the concept of a modified core of Boolean functions with respect to the system of characters of a group on which functions of algebra of logic are given. The criteria of belonging these functions to the class of generalized neurofunctions are provided. The algebraic structure of cores of Boolean neurofunctions is studied. On the basis of properties of tolerance matrices, a number of necessary conditions for implementation of Boolean functions by one generalized neural element are obtained. The obtained results allow to develop efficient methods of synthesis of integer-valued generalized neural elements with a large number of inputs that can be successfully applied in solving information compression and transmission problems, as well as discrete signal recognition problems.     

Refinement algebra with dual operator

Algebras of imperative programming languages have been successful in reasoning about programs. In general an algebra of programs is an algebraic structure with programs as elements and with program compositions (sequential composition, choice, skip) as algebra operations. Various versions of these algebras were introduced to model partial correctness, total correctness, refinement, demonic choice, and other aspects. We introduce here an algebra which can be used to model total correctness, refinement, demonic and angelic choice. The basic model of our algebra are monotonic Boolean transformers (monotonic functions from a Boolean algebra to itself).     

Torsion elements in effect algebras

We define the torsion element in effect algebras and use it to characterize MV-effect algebra and 0-homogeneous effect algebras in chain-complete effect algebras. As an application, we prove that every element of an orthocomplete homogeneous atomic effect algebra has a unique basic decomposition into a sum of a sharp element and unsharp multiples of atoms. Further, we characterize homogeneity by the set of all sharp elements in orthocomplete atomic effect algebras.     

Pseudocomplemented lattice effect algebras and existence of states

We prove that in every pseudocomplemented atomic lattice effect algebra the subset of all pseudocomplements is a Boolean algebra including the set of sharp elements as a subalgebra. As an application, we show families of effect algebras for which the existence of a pseudocomplementation implies the existence of states. These states can be obtained by smearing of states existing on the Boolean algebra of sharp elements.     

Blocks of pseudo-effect algebras with the Riesz interpolation property

 Pseudo-effect algebras are partial algebras (E;+,0,1) with a partially defined sum + which is not necessary commutative only associative and with two complements, left and right ones. They are a generalization of effect algebras and of orthomodular posets as well as of (pseudo) MV-algebras. We define three kinds of compatibilities of elements and we show that if a pseudo-effect algebra satisfies the Riesz interpolation property, and another natural condition, then every maximal set of strongly compatible elements, called a block, is a pseudo MV-subalgebra, and the pseudo-effect algebra can be covered by blocks. Blocks correspond to Boolean subalgebras of orthomodular posets. Dedicated to Prof. Ján Jakubík on the occasion of his 80th birthday The paper has been supported by the grant VEGA 2/3163/23 SAV, Bratislava, Slovakia, and the fellowship of the Alexander von Humboldt Foundation, Bonn, Germany. The author is thankful the Alexander von Humboldt Foundation for organizing his stay at University of Ulm, Ulm, summer 2001, and Prof. G. Kalmbach H.E. for her cordial hospitality and discussions.     

Boolean vector algebraic structures: properties and applications to digital logic simulation

Boolean algebra is the basic mathematic used in the analysis and synthesis of binary systems. Technological advances have led to an increasing interest in multivalued logic systems where more than two logical values are used.In the application of multivalued logic, each logical value could often be represented by a Boolean vector, ie a vector with binary components (0 or 1). Therefore, it is quite important to have a thorough understanding of properties embedded in the algebraic structures using Boolean vectors as basic operands. This gives the motivation for this study.Boolean vector operations are introduced and two major modes associated with the complement are characterized. Then we define a Boolean vector E-algebra and its major features are given. Finally, some applications are discussed and illustrated.     

Boolean lifting property for residuated lattices

In this paper we define the Boolean lifting property (BLP) for residuated lattices to be the property that all Boolean elements can be lifted modulo every filter, and study residuated lattices with BLP. Boolean algebras, chains, local and hyperarchimedean residuated lattices have BLP. BLP behaves interestingly in direct products and involutive residuated lattices, and it is closely related to arithmetic properties involving Boolean elements, nilpotent elements and elements of the radical. When BLP is present, strong representation theorems for semilocal and maximal residuated lattices hold.     

基于目标演绎距离的一阶逻辑子句集预处理方法

一阶逻辑定理证明是人工智能的核心基础,研究一阶逻辑自动定理证明器的相关理论和高效的算法实现具有重要的学术意义。当前一阶逻辑自动定理证明器首先通过子句集预处理约简子句集规模,然后通过演绎方法对定理进行判定。现有的应用于证明器中的子句集预处理方法普遍只从与目标子句项符号相关性角度出发,不能很好地从文字的互补对关系中体现子句间的演绎。为了在子句集预处理时从演绎的角度刻画子句间的关系,定义了目标演绎距离的概念并给出了计算方法,提出了一种基于目标演绎距离的一阶逻辑子句集预处理方法。首先对原始子句集进行包含冗余子句约简并应用纯文字删除规则,然后根据目标子句计算剩余子句集中的文字目标演绎距离、子句目标演绎距离,并最终通过设定子句演绎距离阈值来实现对子句集的进一步预处理。将该预处理方法应用于顶尖证明器Vampire,以2017年国际一阶逻辑自动定理证明器标准一阶逻辑问题组竞赛例为测试对象,在标准的300 s内,加入提出的子句集预处理方法的Vampire4.1相比原始的Vampire4.1多证明4个定理,能证明10个Vampire4.1未证明的定理,占其未证明定理总数的13.5%;在证明的定理中,提出的子句集预处理方法能对77.2%的子句集进行约简,最大子句集约简规模达到51.7%。实验结果表明,提出的一阶逻辑子句集预处理方法是一种有效的方法,能有效地约简一阶逻辑子句集的规模,提高一阶逻辑自动定理证明器的证明能力。     

Extension of belief functions to infinite-valued events

We generalise belief functions to many-valued events which are represented by elements of Lindenbaum algebra of infinite-valued ?ukasiewicz propositional logic. Our approach is based on mass assignments used in the Dempster–Shafer theory of evidence. A generalised belief function is totally monotone and it has Choquet integral representation with respect to a unique belief measure on Boolean events.     

Algebraic models of deviant modal operators based on de Morgan and Kleene lattices

An algebraic model of a kind of modal extension of de Morgan logic is described under the name MDS5 algebra. The main properties of this algebra can be summarized as follows: (1) it is based on a de Morgan lattice, rather than a Boolean algebra; (2) a modal necessity operator that satisfies the axioms N, K, T, and 5 (and as a consequence also B and 4) of modal logic is introduced; it allows one to introduce a modal possibility by the usual combination of necessity operation and de Morgan negation; (3) the necessity operator satisfies a distributivity principle over joins. The latter property cannot be meaningfully added to the standard Boolean algebraic models of S5 modal logic, since in this Boolean context both modalities collapse in the identity mapping. The consistency of this algebraic model is proved, showing that usual fuzzy set theory on a universe U can be equipped with a MDS5 structure that satisfies all the above points (1)-(3), without the trivialization of the modalities to the identity mapping. Further, the relationship between this new algebra and Heyting-Wajsberg algebras is investigated. Finally, the question of the role of these deviant modalities, as opposed to the usual non-distributive ones, in the scope of knowledge representation and approximation spaces is discussed.     

正则FI-代数的刻画及成为Boole代数的条件

正则FI-代数是仅基于蕴涵算子在一般集合上建立的逻辑代数。基于正则FI-代数的公理组以及诸多性质之间的内部联系,给出了正则FI-代数的两个公理组条件更少的刻画定理,简化了正则FI-代数的定义形式。在正则FI-代数中引入蕴涵分配性,探讨了蕴涵分配正则FI-代数的若干性质,证明了蕴涵分配正则FI-代数与Boole代数是相互等价的代数系统,给出了Boole代数的一种新的刻画,使其在形式上更接近于二值逻辑代数。     

命题逻辑提升到一阶逻辑上的子句消去方法

在基于命题逻辑的可满足性问题（SAT）求解器和基于一阶逻辑的定理证明器上，子句集简化一直是必不可少的步骤，而其中子句消去方法在这些子句集简化方法中是非常重要的组成部分。将命题逻辑中的子句消去方法归结隐藏恒真消去方法（RHTE）和归结隐藏包含消去方法（RHSE）提升到一阶逻辑上，并且利用蕴含模归结原则（IMR）证明了这种提升方式在一阶逻辑上具有可靠性（Soundness），即依据这两种子句消去方法删除一阶逻辑公式集中的子句，并不会改变公式集的可满足性或者不可满足性。此外，将这两个方法与一阶逻辑子句消去方法锁子句消去方法（BCE）和归结包含消去方法（RSE）进行组合推广，发展得到一阶逻辑上新型子句消去方法（BC+RHS）E、（RS+RHT）E和（RHS+RHT）E，并且证明了这3种子句消去方法在一阶逻辑上的可靠性。最后，分析比较了这些子句消去方法的有效性，并且证明了这3种新型子句消去方法比组成它们的原始子句消去方法均具有更高的有效性。     

泛逻辑学中UB代数系统的fuzzy滤子

文献[2]给出了理想状态下泛逻辑学的形式演绎系统!,证明了此系统是可靠的。文献[3]提出了在理想状态(h=k=0.5)下泛逻辑学对应的代数系统-UB代数。文献[5]研究了泛逻辑学中UB代数系统的若干性质。在文献[3,5]的基础上,进一步讨论了UB代数fuzzy滤子与商代数。     

对称*-λ-半环

Kleene代数在计算机科学中具有基础而特殊的重要性。在计算机工程应用中,Kleene代数及相关*-半环已被成功应用于基础安全分析、底层程序变换以及并行控制等许多领域。论文给出了对称*-λ-半环的定义及其等价刻画,并指出对称*-λ-半环是Kleene代数概念的推广。     

基于条件事件代数的贝叶斯网的逻辑推理

条件事件代数理论在数据融合系统中有着重要的应用前景,该理论可用来解决不确定性、概率性和模糊性推理问题。条件事件代数是在确保规则与条件概率相容的前提下,把布尔代数上的逻辑运算推广到条件事件(规则)集合中的逻辑代数系统。对于一些特殊的贝叶斯网(如多树型网络)已经有了一些可行的概率推理的算法,但到目前为止,还没有可行的逻辑推理的算法。随着对不确定性知识研究的深入,迫切需要具有逻辑推理的算法。论文介绍了乘积空间条件事件代数的定义和基本性质,提出了基于乘积空间条件事件代数的贝叶斯网的逻辑推理的算法以及应用。     

ON SEMANTICS OF L-VALUED FIRST-ORDER LOGIC Lvft

Abstract

Many-valued logic system always plays a crucial role in artificial intelligence. Many researchers have paid considerable attention to lattice-valued logic with truth values in a lattice. In this paper, based on lattice implication algebras introduced by Xu (Journal of Southwest Jiaolong University (in Chinese), Sum. No. 89(1), 20-27, 1993, and L-valued propositional logic _vft, established by Xu et al. (Information Sciences, 114, 20S-235, 1999a), the semantics of a L-type lattice-valued first-order logic L_vft, with truth values in lattice implication algebras were investigated. Some basic concepts about semantics of L_vftsuch as the language and the interpretation etc. were given and some semantic properties also were discussed. Finally, a concept of g-Skolem standard form was introduced, and it was shown that the unsatisfiability of a given lattice-valued formula was equivalent to that of its g-Skolem standard form. It will become a foundation to investigate the resolution principle based on first-order logic L_vft     

逻辑系统MTL▽及其完备性

在逻辑系统MTL（Monoidal t-norm based Logic）中引入一元逻辑连接词▽,并在原有公理模式的基础上添加若干新公理,构建了一种新的MTL的扩张MTL_▽逻辑系统。为了进行相应的语义研究,在MTL代数的基础上,引入一元算子▽,提出MTL_▽代数的概念,并证明了MTL_▽代数的同构定理。基于线性MTL_▽代数证明了逻辑系统MTL_▽的完备性。     

基于聚类和划分的SAT分治判定

提出了一种将布尔公式划分为子句组来进行布尔可满足性判定的方法.CNF(conjunctive normal form)公式是可满足的当且仅当划分产生的每个子句组都是可满足的,因此,通过判定子句组的可满足性来判定原公式的可满足性,相当于用分治法将复杂问题分解为多个子问题来求解.这种分治判定方法一方面降低了原公式的可满足性判定复杂度;另一方面,由于子句组的判定可以并行,因而判定速度能够得到进一步的提高.对于不能直接产生布尔子句组划分的情形,提出了一种利用聚类技术将CNF公式聚类成多个簇,然后消去簇间的公共变量来产生子句组划分的方法.     

由可精确测量元控制的弱可换的伪效应代数

主要研究了由可精确测量元控制的弱可换的伪效应代数中可精确测量元。证明了可精确测量元控制的弱可换的伪效应代数中可精确测量元是弱可换的伪正交代数代数。讨论了弱可换的伪效应代数与BZ-偏序集之间的关系。讨论了弱可换的伪效应代数商代数中可精确测量元与正规Riesz理想之间的关系。