Author(s): Tarek Merabtene
Linked Author(s):
Keywords: Flood risk management; Retarding basin control; Floodgates operation; Fuzzy logic
Abstract: For emergency and disaster control of flood levels in retarding basins, optimal operation of the in situ emergency floodgates is a major concern of the river engineers in charge. Past flood records and operation practices proved that engineers with different prior experience are expected to behave differently during a flood event. Among all decision variables the forecast accuracy of water levels in the retarding basins and the river channel concurrently remains the most critical. Uncertainties are driven by the prediction of the stochastic nature of inflows from the main river and its tributaries to the retarding basins and the forecast accuracy of rainfall. This paper presents the application of combined fuzzy sets to improve water levels forecast in complex multi-retarding basins system. The forecast model integrates fuzzy pattern matching for rainfall forecast module as feedforward to the fuzzy inference module for water level forecasting. Furthermore, to overcome some of the existing complexities involved in the operational management of emergency floodgates a fuzzy controller decision support system (FC-DSS) has been developed. In addition to the above modules, the FC-DSS integrates past knowledge of experienced river engineers and a 2D runoff model. The procedure has been applied to real data of the Tsuboi river flood control system and proved to be very efficient and practically robust to improve the operational management of the flood emergency gates.
Year: 2017