Manchester Metropolitan University's Research Repository

Development of an intelligent sensing technique for active control of railway vehicles with independently rotating wheels

Derbyshire, Daniel (2015) Development of an intelligent sensing technique for active control of railway vehicles with independently rotating wheels. Masters thesis (MPhil), Manchester Metropolitan University.


Available under License Creative Commons Attribution Non-commercial No Derivatives.

Download (1MB) | Preview


This thesis continues the development of an intelligent sensing scheme; using practical techniques for estimating the vehicle variables and economic measurements, which are mounted away from any arduous environments. The independently rotating wheelset design (IRW) de-couples the wheels; losing the mechanical feedback inherent with the conventional solid axle. The issue with the IRW, however, is that the natural curving characteristic is lost as a result, and therefore it is necessary to provide steering to the wheelset, to avoid flange contact and guide the wheelset, which can be supplied using a yaw constraint: passively or actively. As primary feedback variables aren’t readily available to measure, and are costly to provide, the Kalman filter has been used to provide full state optimal estimation of the dynamic system. Involving random perturbations to return the necessary states required, the Kalman filter allows the controller and actuator to apply active steering to the IRW. This thesis applies the Kalman-Bucy filter to a closed loop system with a mechatronic vehicle. A simple P controller is used to provide a torque to the actuators. The novelty about this work is that the sensors are mounted to the vehicle body, avoiding the extreme climactic conditions that a sensor would usually see if mounted to the axle. Re-formulating the Kalman filter to include curvature and cant within the state space and output matrices has been assessed to see if the curving performance will be maintained or improved. Altering the amount of sensors used has also been assessed, to see how the curving performance is affected. This can coincide with assessing whether the system will work, should one or more of the sensors fail.

Impact and Reach


Activity Overview

Additional statistics for this dataset are available via IRStats2.

Actions (login required)

View Item View Item