Adriana Maggiore

FROS: a fuzzy logic-based recogniser of olfactory signals

Abstract In this paper we describe FROS, a fuzzy logic-based recogniser of olfactory signals. FROS integrates two recognisers, namely the shape-based recogniser and the dynamic range-based recogniser. While the former uses a linguistic description of the shape of the signals, the latter exploits a fuzzy classification of their dynamic ranges. FROS was designed to classify signals produced by a sensor array that comprises conducting polymer sensors with partially overlapping sensitivities. The sensors are exposed to odorants and the resistance values are used for classification. Results of the application of FROS to two different test cases are also presented.

Mixing fuzzy, neural and genetic algorithms in an integrated design environment for intelligent controllers

In this paper the concept of hierarchical hybrid fuzzy controllers (HHFCs) is introduced. By means of HHFCs, the development of neuro-fuzzy controllers for complex plants can be simplified, just by dividing the controller into blocks, each of which solves a particular subset of control tasks.

An integrated hybrid approach to the design of high-performance intelligent controllers

This paper presents a hybrid approach to the development of high-performance real-time intelligent and adaptive controllers for nonlinear plants. Several paradigms derived from cognitive sciences ore considered and analyzed in this work, such as neural networks, fuzzy inference systems, genetic algorithms, etc. Although most of these paradigms are widely known and have been used extensively in the field of automatic control since several years, the novelty of the proposed approach resides in their tight integration and its capability of allowing a hybrid design. The different control strategies have also been integrated with the theory of finite state automata, in such a way that an automaton tracks the different plant states and selects accordingly one out of a given number of controller characteristics, each one being designed in a hybrid manner. State transitions can also be triggered by fuzzy and neural signals. Finally, two practical examples of the proposed hybrid approach are analyzed.

A new linguistic fuzzy approach to recognition of olfactory signals

In this paper we propose a new fuzzy linguistic method to recognize odorant samples. The method is applied to raw experimental data collected from a sensor array that comprises sixteen conducting polymer sensors with partially overlapping sensitivities. The sensors are exposed to odor samples and the percentage change in resistance is used for classification. The method describes the shape of each sensor response in terms of linguistic expressions derived from a fuzzy partition of the area occupied by the response. A purposely-defined weighted distance is used to compare the linguistic descriptions. Results on the application of the method to the classification of three chemicals in two different concentrations are presented.

EL-SIM: a Development Environment for Neuro-Fuzzy Intelligent Controllers

This paper presents a new technique for the design of real-time controllers based on a hybrid approach which integrates several control strategies, such as intelligent controllers (e.g., artificial neural networks, fuzzy systems), traditional linear controllers, finite state automata. An integrated programming environment, called EL-SIM, is also presented, suited for developing high-performance intelligent controllers for industrial applications. EL-SIM provides general tools to support the development and optimization of control systems based on the aforementioned approach, by means of several cognitive or hybrid algorithms, which allow also improvement of environmental performance indexes, like power consumption or toxic waste emission. EL-SIM permits both the study of new experimental techniques in research application and the design, tuning and testing of widely used control architectures for industrial applications.

Linguistic modelling based on experimental data

This paper describes a method for constructing linguistic models from observed data. A linguistic model is derived from the reduction, based on clustering, of the number of fuzzy sets and rules which constitute a fuzzy model. This, in its turn, is built by applying a new method, called the local approximation method which determines a piecewise linear approximation of a set of samples of the system to be modelled. The approximation error due to linearisation can be chosen based on the degree of detail of the required model. In particular, if the final model is a linguistic one, the major requirements are readability and understandability, which normally correspond to reduced precision.