Physical tidal models are used to investigate the complex coastal and estuarine environments. These models can be used to aid in the design of a hydraulic structure which may affect, or be affected by, the complex flow patterns and bed morphology found within such regions. This thesis describes an investigation into the control of the dynamic water-level within a physical tidal model, manipulated using a pneumatic tide generator. A digital control system is described which samples the model depth and uses this information to adjust the state of a valve actuator. A butterfly valve has traditionally been used for this purpose, but a new needle valve is proposed which has mechanical and economic advantages. A control algorithm is used to dictate the state of the actuator to ensure that the model dpeth is a as close to the demand as possible. Several algorithms have been investigated; proportional plus integral, integral only, predictive and gain scheduling. The results obtained using this control system show that good control of the model water level can be achieved. This is dependant, however, upon several factors; the model depth has to be measured accurately, the rate of airflow into the pneumatic tank has to be adjusted very finely and the control system has to be tuned properly. It has been concluded that the integral only algorithm is most suitable for the system studied here, because only one tuning coefficient has to be selected. The success of a control system is measured by how well the resulting tidal patterns follow the demand. The difference, or error, is analysed in both the time and frequency domains. Several error types have been pin - pointed. These include phase shift, harmonic distortion, overshoot and higher frequency components.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:637307 |
| Date | January 1993 |
| Creators | Horn, C. D. |
| Publisher | Swansea University |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
Page generated in 0.0017 seconds