Vibrations control of light rail transport systems

This doctoral thesis mainly focuses on ?the hunting oscillations? in which the lateral and yaw movements are seen together which is the natural consequence of the dynamics of wheel-rail contact, that arises on a straight road independently from the disturbances due to the wheel-track geometry. Although hunting oscillations decrease the wear problem in the wheel and rail, they lead to flange climb and then derail by increased amplitudes of the oscillations at high speeds. This condition causes system instability and a limitation in the vehicle speed on straight and smooth rail, even if there is no any disorder on the line or on the rails of the systems. Instability of the system caused by hunting oscillations is a significant problem in light rail system as in all rail systems. Increasing operating speeds of light rail systems is a basic need because of increasing distances between stations and the need for raising the passenger transportation capacity in major cities such as Istanbul. Within this context, a domestic manufacturing light rail vehicle which is used in Istanbul transportation is discussed and various analyses are performed to examine the hunting oscillations mainly. In this thesis, firstly wheel ? rail interaction is linearized by using a single wheelset model. Following that, light rail vehicle models are developed to deal with lateral oscillations mainly. These models are established as 27 degrees of freedom in Matlab that deals with only lateral amplitudes and 54 degrees of freedom in Simpack that deals with all movements. Natural frequencies of these models are obtained by modal analysis, and then the results are presented and compared with each other. Similarly, displacement and acceleration responses are obtained in time domain under the effect of lateral disturbances. Also, frequency responses of these models which are obtained by two different software are compared. Thus, these models are verified. After this stage, critical speeds of the light rail vehicle are identified with the help of stability analysis by using the Matlab models. It is investigated that, how the critical speed of a light rail vehicle is affected by changing equivalent conicity (lambda), wheel radius (r0) and stiffness of primary and secondary suspensions which are parameters that influence the critical velocity of rail vehicle. Afterwards, primary and secondary stiffness are optimized by Matlab Genetic Algorithm toolbox. The results of improvement by the optimization are presented with the results of analysis that are performed in time and frequency domain. Furthermore, critical speed is determined by looking at the relationship of the rail vehicle speed with the frequency response of the system. This method has never been used in the literature and it is used for the first time in this thesis. Then, in order to increase the critical speed, lateral and yaw movements of the wheelsets are controlled by using actuators which are placed between wheelsets and bogies in lateral direction. LQR (Linear Quadratic Regulator) algorithm is used as an optimal state feedback control algorithm to control the movements of wheelsets. The system responses are analyzed in different cases such as using existing suspensions, optimized suspensions and controlled wheelsets. Then, the results are presented with comparison. Hunting oscillations are suppressed to guarantee very high speeds with the control application. All analysis in this thesis is performed for the no-load and full- load cases of light rail vehicle. Thus, the stability improvements applied by the change of the mass and inertia values of light rail vehicle are presented. Moreover, in this thesis, a conventional PID type fuzzy controller and parameter adaptive fuzzy controller are designed to control vibrations actively of a light rail transport vehicle which modeled as 6 degree-of-freedom system and compared performances of these two controllers. To control vibrations actively, a PID type fuzzy controller which is obtained by combining fuzzy PI and fuzzy PD controllers is preferred because of its robust character and superior performance. The PID type fuzzy controller using parameter adaptive method is designed by tuning the parameters online. In order to obtain higher performance from this controller is studied. As a result of this thesis, the vibration analyses of a domestic production light rail vehicle used in Istanbul transportation are carried out and the stability of this light rail vehicle is examined in terms of hunting oscillations. Then, the lateral stability and critical velocity of the light rail vehicle are increased by the use of optimization and control applications to achieve running safety. Also, passenger comfort is increased by controlling vertical vibrations of this vehicle.