An Autonomic Methodology for Embedding Self-tuning Competence in Online Traffic Control Systems

Recent advances in technology, control and computer science play a key role towards the design and deployment of the next generation of intelligent transportation systems (ITS). The architecture of such complex systems is crucial to include supporting algorithms that can embody autonomic properties within the existing ITS strategies. This chapter presents a recently developed adaptive optimization algorithm that combines methodologies from the fields of traffic engineering, automatic control, optimization and machine learning in order to embed self-tuning properties in traffic control systems. The derived adaptive fine-tuning (AFT) algorithm comprises an autonomic tool that can be used in online ITS applications of various types, in order to optimize their performance by automatically fine-tuning the system’s design parameters. The algorithm has been evaluated in simulation experiments, examining its ability and efficiency to fine-tune in real time the design parameters of a number of traffic control systems, including signal control for urban road networks. Field results are in progress for the urban network of Chania, Greece, as well as for energy-efficient building control. Some promising preliminary field results for the traffic control problem of Chania are presented here.

[1]  Mokhtar S. Bazaraa,et al.  Nonlinear Programming: Theory and Algorithms , 1993 .

[2]  Markos Papageorgiou,et al.  A Multivariable Regulator Approach to Traffic-Responsive Network-Wide Signal Control , 2000 .

[3]  Markos Papageorgiou,et al.  Adaptive Fine-Tuning of Nonlinear Control Systems With Application to the Urban Traffic Control Strategy TUC , 2007, IEEE Transactions on Control Systems Technology.

[4]  M. Papageorgiou,et al.  Adaptive fine-tuning of non-linear control systems with application to the urban traffic control strategy TUC , 2007, 2007 European Control Conference (ECC).

[5]  Chee Kheong Siew,et al.  Universal Approximation using Incremental Constructive Feedforward Networks with Random Hidden Nodes , 2006, IEEE Transactions on Neural Networks.

[6]  Yu. M. Ermol’ev.  On the method of generalized stochastic gradients and quasi-Féjer sequences , 1969 .

[7]  J. Kiefer,et al.  Stochastic Estimation of the Maximum of a Regression Function , 1952 .

[8]  Elias B. Kosmatopoulos,et al.  Large Scale Nonlinear Control System Fine-Tuning Through Learning , 2009, IEEE Transactions on Neural Networks.

[9]  J. Spall.  Multivariate stochastic approximation using a simultaneous perturbation gradient approximation , 1992 .

[10]  H. Robbins.  A Stochastic Approximation Method , 1951 .

[11]  Marios M. Polycarpou,et al.  High-order neural network structures for identification of dynamical systems , 1995, IEEE Trans. Neural Networks.

[12]  Markos Papageorgiou,et al.  Adaptive Performance Optimization for Large-Scale Traffic Control Systems , 2009, CTS 2009.

[13]  Martin A. Riedmiller,et al.  RPROP - A Fast Adaptive Learning Algorithm , 1992 .

[14]  Anastasios Kouvelas.  Adaptive fine-tuning for large-scale nonlinear traffic control systems , 2011 .

[15]  Elias B. Kosmatopoulos,et al.  International comparative field evaluation of a traffic-responsive signal control strategy in three cities. , 2006 .

[16]  Markos Papageorgiou,et al.  Extensions and New Applications of the Traffic-Responsive Urban Control Strategy: Coordinated Signal Control for Urban Networks , 2003 .