Guest Editorial

摘要

Over the past few years, the development of intelligent decision-making techniques for complex systems has been significantly advanced by using fuzzy logic and neural network based methods.The application of fuzzy logic and neural networks has captured two of the essential features in human decision-making, i.e., the linguistic nature of knowledge representation, which facilitates the process of knowledge acquisition and transfer, and the learning nature of knowledge evolution, which leads to improvement in system performance and knowledge.This is due to the fact that, in fuzzy logic, linguistic variables are used as basic terms for knowledge descriptions, while in neural networks, learning algorithms are employed as the basic mechanism for universal approximation.This special issue is dedicated to the applications of fuzzy logic and neural network techniques for solving real-world problems.Its purpose is to collect and disseminate some focused results of ongoing research and development efforts in these two methods, especially in their combinations.The decision to organize this special issue was made at the Fourth International Symposium on Robotics and Manufacturing (ISRAM '92) in Sante Fe, New Mexico.A call for papers resulted in 15 responses, from which the review process selected 7 for publication.The papers can broadly be classified into four categories: applications in robotic systems, industrial applications, inductive reasoning for fuzzy system design, and combination of fuzzy logic and neural networks.The first paper, by Cao and Sanderson, introduces the concept of generalized fuzzy Petri nets (FPNs) and the reasoning structures of their transitions.The proposed FPN model uses three types of fuzzy variables, called local fuzzy variables, fuzzy marking variables, and global fuzzy variables, respectively, to describe uncertainty based on different aspects of fuzzy information.Analytical properties of several basic types of FPNs are presented in the paper.