On the Inner Product Operation of Discrete Z-Numbers

Ruolan Cheng,Yong Deng

DOI: https://doi.org/10.1109/ICUS58632.2023.10318439

2023-01-01

Abstract:Classical linear algebra and fuzzy linear algebra have been extensively studied. However, they are limited since they can only be applied to accurate or fuzzy numbers, which are insufficient to describe real-world information's imperfection and partial reliability. In light of this, Zadeh proposed a new framework for modeling real-world uncertain information: Z-number. The inner product of Z-numbers is an important mathematical tool for reasoning and analyzing Z-number data. However, the existing Z-number inner product operation method does not follow Zadeh's expansion principle but only corresponds to the scalar value. To overcome the shortcomings and further develop the Z-number linear algebra space, a new Z-number inner product operation is proposed based on the classic fuzzy inner product algorithm and Zadeh's principle of extension. This paper aims to provide a new direction for the inner product of Z-numbers that aligns better with the original meaning of Z-number. The proposed method obtains the optimal underlying probability distribution of Z-number through a genetic algorithm-based maximum entropy optimization model and then uses the basically uniform fusion of the probability distribution proposed by Yager to process it. Finally, a numerical example illustrates the inner product operation process of two Z-numbers, and a prospect for the future research direction is made.