Journal of Hydrologic Engineering · 2005 · 333 citations · 21 references
Natural ResourcesComputer ScienceHydrologyArtificial IntelligenceForecastingMachine LearningFlood ForecastingFlood WarningHydrologic Time Series AnalysisAccuracyFuzzy Inference SystemsHybrid AlgorithmsHybrid Evolutionary AlgorithmsLinear RegressionMachine Learning-based Hydrologic ForecastingNeural ForecastingRiver Flood Forecasting
In a flood‑prone region, quick and accurate flood forecasting is imperative, extending lead time for disaster warnings and allowing residents to evacuate. The study evaluates two hybrid AI models—ANN‑GA and ANFIS—for flood forecasting along a Yangtze River channel. Performance is benchmarked against an empirical linear regression model, with downstream Han‑Kou water levels forecasted from upstream Luo‑Shan levels. Both hybrid models outperform the linear regression, with ANFIS achieving the best accuracy yet requiring many parameters, while ANN‑GA also performs well but needs longer computation time and extra parameters.
In a flood-prone region, quick and accurate flood forecasting is imperative. It can extend the lead time for issuing disaster warnings and allow sufficient time for habitants in hazardous areas to take appropriate action, such as evacuation. In this paper, two hybrid models based on recent artificial intelligence technology, namely, the genetic algorithm-based artificial neural network (ANN-GA) and the adaptive-network-based fuzzy inference system (ANFIS), are employed for flood forecasting in a channel reach of the Yangtze River in China. An empirical linear regression model is used as the benchmark for comparison of their performances. Water levels at a downstream station, Han-Kou, are forecasted by using known water levels at the upstream station, Luo-Shan. When cautious treatment is made to avoid overfitting, both hybrid algorithms produce better accuracy in performance than the linear regression model. The ANFIS model is found to be optimal, but it entails a large number of parameters. The performance of the ANN-GA model is also good, yet it requires longer computation time and additional modeling parameters.
21