Tidal barrage power plants leverage the daily variations in sea water level to generate electricity. Optimising tidal barrage operation requires the solution of an associated continuous-time optimal control problem, by looking at the past tidal elevation, as well as the predicted tidal elevation during subsequent tidal cycles. Although highly predictable, the tidal level variations are also influenced by stochastic weather conditions, which are often neglected when assessing tidal barrage energy generation. In this study, a receding-horizon control framework is proposed, whereby optimal operation of tidal barrages can be computed, while including both past observations of the tidal elevation, as well as a tidal level forecast, based either on weather data or on the astronomic tide alone. At each step of the receding-horizon algorithm, tidal barrage optimal operation is computed using moment-based control, a mathematical method that enables the transcription of the infinite-dimensional tidal barrage optimal control problem onto a finite dimensional nonlinear program. With this framework, the impact of considering the storm surge within the control calculations, as opposed to only considering astronomic tidal forcing, is assessed. Including a state-of-the-art weather-informed forecast, as opposed to a purely astronomic forecast, in the representation of the tidal elevation is shown to improve the performance of the controller, highlighting the value of accounting for the storm surge when computing tidal barrage optimal operation. • The impact of storm surge on tidal barrage optimal control is studied. • Receding-horizon control enables including real observations as well as a tidal forecast. • Using a weather-informed tidal forecast is shown to enhance controller performance.
Skiarski et al. (Fri,) studied this question.