Enhanced safety and reliability for trains through fault tolerant control

EPSRC/RRUK grant, 10/2009 -09/2012. Principle Investigator – Professor T X Mei

Active controls for railway vehicles can provide much better performance, improve operation efficiency and reduce track damages. However the introduction of additional control system and associated electronics into the traditional mechanical system will have to meet the demanding requirements for safety and reliability in rail industry.

This project seeks to investigate fault tolerant control methods that will be able to ensure the safety and improve the overall reliability of rail vehicles, paving the way for an increased industrial acceptance and a wider application of advanced active systems. A key research objective is to develop practical solutions that do not demand a high level of redundancy and hence reduce the system costs.

The project is a natural extension of the leading research work at Salford University in the areas of high-integrity actuation and fault detection methods.

This project is funded by EPSRC (UK) via Rail Research UK grant.

Novel integrated control for vehicle traction and steering systems

(TSB KTP Grant, 11/2007-11/2010. PI – Professor T X Mei, in collaboration with GGS Engineering)

The increased awareness for the need for low carbon based transport has helped to accelerate the development of both public transport vehicles and personal cars. It is expected that the next generation of the vehicles must not only be energy efficient and produce less (direct and indirect) pollution, but also improve their attractiveness to the public through for example enhanced performance, safety, comfort and ease of use etc.

This project is in collaboration with GGS Engineering in Derby to develop a high performance and ultra efficient control system for the integrated provision of traction and steering with the use of in-wheel electric motors, i.e. motors built into wheels. The use of the wheel motors offers advantages of mechanical simplicity and power loss reduction, but this study also seeks to control the electric motors on the two sides of a vehicle independently for the provision of vehicle steering/guidance in a new form of steer-by-wire. The technology developed in this project will be suitable for applications in both railway and automotive industries.

This project is jointly funded by TSB (Technology Strategy Board, UK) through the Knowledge Transfer Partnership Scheme and the industrial partner GGS Engineering.

Test Vehicle

Measurement of vehicle ground speed by exploring vehicle dynamics with intelligent data processing

EPSRC/RRUK grant, 01/03/2009-31/12/2009. PI – Professor T X Mei

The speed of a vehicle is normally measured by sensing the rotational speed of the wheels through the use of slotted discs mounted onto the wheels or in railway vehicles through detecting the number of teeth of a traction gear box. As a standard practice for many years, this conventional approach is considered to provide sufficient accuracy as long as there is no excessive wheel slip or slide – a phenomenon that causes wheels to rotate much faster or lower than the equivalent of the vehicle.

This project carries out a more detailed study into a novel concept, originated by the PI at Salford University (Professor T X Mei), and the overall aim of the proposal is to establish an important foundation for the study of a reliable and low cost technique for the measurement of train ground speed that will not be compromised in the wheel slip or slide conditions. The new method derives the train speed from time shifts between the vertical motions of two wheels of a vehicle/bogie, which are in turn estimated using measurement signals from two inertial sensors mounted on a vehicle/bogie frame. This may be further enhanced with data fusion with wheel based encoders to reduce measurement delays at extremely low speeds and to improve the measurement robustness. 

This project is funded by EPSRC (UK) via Rail Research UK grant.

Principle of the new measurement method

System integration for rail wheelset steering and traction control

EPSRC grant, 02/2002-08/2003. Principle Investigator – Professor T X Mei

The suspension and guidance systems for the control of vehicle dynamics deal with issues such as vehicle stability, running behaviour and passenger ride comfort. These have been traditionally achieved through the use of passive components. However, active controls via mechatronic and electronic components can push the boundaries and achieve far beyond what is possible with mechanical suspensions. More recent research has shown that wheelsets can be controlled actively in a manner that the adverse forces at the wheel-rail contact which exist with passive or mechanical solutions are significantly reduced. Extensive studies have concentrated upon individual active components to provide particular functions and their development has been focussed on issues at the sub-system level, e.g. the design conflict between ride quality improvement and actuation requirement for the secondary active suspensions.

This project carries out a feasibility study of a systems approach that will bring together the active steering of wheelset and traction control which interact at the wheel-rail contact points, and it is based upon a long term prospect where active wheelset control will be used for rail vehicles and an integrated control for the two sub-systems will then become inevitable in order to ensure the stability and performance of the system. As the dynamics of a railway vehicle is known to be highly interactive, it is essential to understand properly how the structure dynamics and active functions will affect one another; and what adverse effect the interactions will have on the overall performance of the vehicle. The research has helped to establish a rigorous design methodology for the development of active control for railway wheelsets that are also equipped with traction motors.

This project is funded by UK Research Council EPSRC.

Development of an actively steered bogie

Funded by Bombardier Transportation, 10/2001-01/2004. Principle Investigator – Professor T X Mei

Active control technology with the use of sensors, control electronics and actuators, can deliver the vehicle performance far beyond what is possible with the traditional passive suspensions. The actively controlled tilting trains are now in day-to-day operation, e.g. in Europe and Japan and beyond, to increase the train speed without the need for a new railway infrastructure and active secondary suspensions are also being considered for commercial applications. However the latest research suggests that active control of wheels/wheelsets directly (also known as Steer-by-wire for railways) can provide far better performance gains and potentially revolutionise future railway vehicles.

Built on world leading research of many years in the field by the Principle Investigator (Prof T X Mei), this project is focused on the practical implementation of the latest control strategies developed by the PI for a full size railway vehicle and demonstrates experimentally the benefits of the active solutions – a world first of its kind. This project is a collaboration between two universities (the other university being Loughborough) and the industrial sponsor Bombardier Transportation. The universities were responsible for the development, implementation and testing of the control strategies, whereas the industrial collaborator undertook detailed hardware development for the bogie as well as the performance assessment using their comprehensive computer simulation software.

The project is fully funded by the industrial partner Bombardier Transportation.

Test Vehicle with actively steered bogie on a full size roller rig

Modified bogie for active steering

Re-adhesion control for AC traction systems, PhD project, 2003-06

Supervisor – Professor T X Mei

Wheel slip is a difficult problem in traction control systems which occurs when applied tractive effort exceeds the level of maximum adhesion (friction) available at the wheel-rail interface, e.g. in poor weather conditions or with contaminated tracks. Apart from the potential impact on normal operations of a rail network, the wheel slip/slide causes undesirable wear to both wheel/track surfaces and increases the cost of maintenance. Most conventional wheel slip protection (or re-adhesion control) schemes detect and control the slip ratio (relative speed between a wheel and the train) to be below a pre-tuned limit and in more extremely cases to control the wheel rotational acceleration below a threshold as an additional measure, but there are a number of known performance constraints.

This project investigates a novel mechatronic solution for the slip detection and re-adhesion control of traction systems in railway vehicles, which explores the changes in dynamic behaviour of railway wheelsets as an indirect indicator of contact conditions. The proposed approach offers a simple and effective method to maximise the use of adhesion available at the wheel-rail contact interface which is essential for the safe and reliable railway operations.

The project is funded by UK ORS Award and Tetley & Lipton award

Condition monitoring of nonlinear systems via multiple Kalman filters, PhD project, 2008-11

Supervisor – Professor T X Mei

Model-based methods are widely studied for applications in FDI and general condition monitoring, but are more suited to tackle linear time invariant systems and can be problematic for systems with very complex and non-linear properties. This project studies a multiple Kalman filters approach that divide a non-linear problem into a number of linearised models that represent an approximation of different conditions. A selected number of Kalman filters are then used to assess the likelihood of a ‘best’ match of the condition of the system is operating at any particular instance to achieve the desired condition monitoring.

This project is supported financially by Pakistani Government through a postgraduate scholarship programme.

Intelligent fault detection for dynamic systems, PhD project, 2005-09

Supervisor – Professor T X Mei

On line fault detection and condition monitoring for dynamic systems are becoming increasingly important because of the potential benefits to detect component failures at their early stages, to prevent further deterioration in performance as well as to ensure timely repair/replacement of faulty components. In the long term, the availability of reliable condition monitoring systems can replace scheduled regular services with maintenance on demand - leading to substantial savings in the total life cycle costs.

This project carries out a fundamental study into a completely new approach, originally proposed by the principle investigator, which offers a simple but very effective means for condition monitoring. The principle is highly innovative which focuses on the comparison of dynamic behaviours between different parts of a system where a symmetry exists in structure and/or in the components used (under the normal condition) and explores changes in dynamic interactions caused by the occurrence of an abnormal condition or component failure. Using a railway vehicle as a case study, the study shows that the proposed condition monitoring technique provides a much improved sensitivity to fault(s), compared to more conventional methods, and is much more robust against uncertainties of external condition changes. The principle is now being studied for condition monitoring for other applications.

The project is financially supported by the Centenary Award (for excellent overseas student)

System-on-chip approach for real time simulation, PhD project, 2006-10

Supervisor – Professor T X Mei

The implementation of complex control laws or computationally demanding simulation algorithms for real-time applications often requires the use of high performance processors and sometimes complex multi-processor architectures. In particular, hardware-in-the-loop approaches are now becoming increasingly popular in the development of control solutions for large and complex systems where full scale testing can be time consuming and/or expensive to carry out.

This innovative project is concerned with the development of a system-on-chip based accelerator for the real time simulation of complex and computationally intensive mathematical models. The main aim is to enable real time control implementation/testing in a hardware-in-loop environment, but it can also be used to speed up computer simulations in the study and design of complex systems. The design of a multi-core accelerator with six CPUs and two sets of floating point units is implemented on a single medium sized and low cost FPGA (field programmable gate arrays) device and is shown to deliver a far superior computation performance than high performance DSPs or the latest PCs.

The project is funded by UK ORS Award and Tetley & Lipton award