Vehicle System Dynamics Vol. 44, Supplement, 2006, 834–842 Intelligent control approaches for tilting railway vehicles H. ZAMZURI*,A. C. ZOLOTAS and R. M. GOODALL Systems and Control, Department of Electronic and Electrical Engineering, Loughborough University, Leicestershire, UK Fuzzy logic has proved to be an effective methodology for many control system applications as it mimics human control logic. This paper presents the work on fuzzy control design for achieving improved local-per-vehicle tilt performance. Two fuzzy-based nulling-type tilt control schemes are discussed, a Fuzzy-PI, and a PID with fuzzy correction. The performance of the control schemes is assessed via appropriate computer simulations and a recently proposed tilt control assessment method. Keywords: Railway dynamics; Tilting trains; Tilt control; Intelligent control; Fuzzy logic 1. Introduction Active tilting technology has been widely introduced for high-speed trains over the last 20 years, with most European trains nowadays fitted with tilt technology. The concept of tilt is rather straightforward: reduce the lateral acceleration experienced by the passengers, by leaning the bodies of the vehicles inwards on curves, thereby enabling higher vehicle speed operation. Early tilt control systems employed feedback control from a lateral accelerometer mounted on the body of the vehicle. However, it proved difficult to achieve a sufficiently fast response on curve transitions without causing a deterioration of the ride quality on straight track. Most tilting train implementations now use the precedence control schemes [1], in which a bogie-mounted accelerometer from the vehicle in front is used to provide ‘precedence’, carefully designed so that the delay introduced by the filter compensates for the preview time corresponding to a vehicle length. Such a control scheme is quite complex; among other things it must reconfigure when the train changes direction, and it is also difficult to provide a satisfactory performance for the leading vehicle of the train. Recent work on local-per-vehicle nulling-type tilt control using modern control approaches has been reported [2]. This paper reports on an extension of these ideas using intelligent methods focusing on fuzzy control technology, the overall objective of the research being to discover whether heuristically derived and/or nonlinear control strategies can give improved performance. *Corresponding author. Email: h.zamzuri@lboro.ac.uk Vehicle System Dynamics ISSN 0042-3114 print/ISSN 1744-5159 online © 2006 Taylor & Francis http://www.tandf.co.uk/journals DOI: 10.1080/00423110600886861