As a means of reducing over-design and gaining greater insight into the operation of dynamic systems, probabilistic methods have evolved as a extremely important design tool. This has been true of power system design in general over the past thirty years, and probabilistic methods are now used in the design of overhead line supports, fault-clearing plant and the assessment of load demands. However, the application of probabilistic methods to the design and analysis of traction power systems has been relatively limited. These systems involve not only the electrical characteristics of the power system, but also the mechanical aspects of an electric railway, and produce moving, frequently changing electrical loads. The electrical system itself is not easily analysed, since modern locomotives incorporate power electronic switching drives, conventionally analysed in the time domain, whilst the supply is analysed in the frequency domain. This study details the development of an object-oriented traction system model, which can incorporate both the time and frequency representations into a single integrated model. This integrated model is then used as the basis of investigation into the probabilistic characteristics of the electrical load and harmonic distortion produced by the traction system. The study shows that an object-oriented approach may be applied to switching-circuit analysis, and that a combined frequency- and time- domain analysis may be easily implemented with this type of system. The probabilistic simulations show for the first time that the load and harmonic characteristics of an a.c. traction system do not fit known parametric models, and are in fact highly dependent upon the operational characteristics of the traction system, e.g. route profile and timetable.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:637271 |
| Date | January 1995 |
| Creators | Hewings, D. B. |
| Publisher | Swansea University |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
Page generated in 0.0026 seconds