A non-cooperative neuro-fuzzy system for integrating ATIS and ATMS decisions

• Published in: Fuzzy sets and systems Vol. 144; no. 2; pp. 313 - 343
• Main Author: Hawas, Yaser E.
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 01.06.2004; Elsevier
• Subjects: Applied sciences; Circuit properties; Computer science; control theory; systems; Control theory. Systems; Digital circuits; Electric, optical and optoelectronic circuits; Electronic circuits; Electronics; Exact sciences and technology; Fuzzy logic; Information, signal and communications theory; Integration; Mathematical methods; Miscellaneous; Neural nets; Optimization; Signal control; Simulation; System theory; Telecommunications and information theory; Theoretical computing; Traffic assignment; Fuzzy logic; Signal control; Integration; Neural nets; Simulation; Traffic assignment; Optimization; Fuzzy system; Methodology; Information system; Optimization method; Traffic; Calibration; Fuzzy set; Algorithm; Decision; Implementation; Engineering; Traffic management; Network; Information processing; Development; Integrated system; Robustness
• ISSN: 0165-0114; 1872-6801
• DOI: 10.1016/S0165-0114(03)00122-2

This paper addresses the problem of integrating successful existing implementations of Advanced Traveler Information Systems (ATIS) and Advanced Traffic Management Systems (ATMS). The methodologies and challenges to integrate ATIS and ATMS are addressed in details. This paper discusses in details the development of a rule-based neuro-fuzzy logic to integrate existing ATIS and ATMS when they operate in a non-cooperative manner. The individual existing logics are initially upgraded to the so-called ATIS and ATMS stand-alone systems, which may be regarded as bi-level optimization systems. The upper level (of the bi-level optimization system) represents an augmented process for guessing the counter system's decisions, and the lower level represents the existing logic (of ATIS or ATMS). The stand-alone systems are then replicated by simulation-based optimization algorithms, which are used to generate the training data necessary to calibrate the neuro-fuzzy logic. The role of the neural nets and the methodology of fuzzy-logic calibration are discussed in details. The effectiveness and robustness (of the neuro-fuzzy integrated ATIS/ATMS system) are assessed using simulation-based experiments with two different hypothetical networks.