Approximate reasoning based on linguistic truth value with alpha-operator

In this paper we discuss the approximate reasoning based on linguistic trut h value (RLV) with the alpha-operator. A new method of RLV is proposed. It is equivalent to the original RLV method, but without using the linguistic truth value fuzzy set. Under given conditions, we convert the problem of ca lculating sup(a' boolean AND (a alpha b)) into the problem of solving the e quation of a'Ta = b. This significantly simplifies the computations of RLV. Some properties of the RLV are studied with examples, such as the "fuzzine ss" of the RLV based on deferent alpha-operators. A graphical method of RLV is also given to provide a convenient tool to implement the proposed metho d. (C) 1999 Elsevier Science B.V. All rights reserved.