FLOOD ESTIMATION METHOD USING MAPLE FORECASTED PRECIPITATION DATA AND APPARATUS THEREOF

摘要

The present invention relates to a flood forecast method using MAPLE weather forecast data and a device thereof. According to the present invention, the flood forecast method using MAPLE weather forecast data comprises a step of extracting hourly MAPLE rainfall forecast data for a flood forecast area from a weather radar; a step of extracting hourly observed mean rainfall volume data for the area from a rainfall observatory in the area; a step of extracting hourly runoff volume data corresponding to hourly water level observation data from a water level observatory in the area; a step of forming flood forecast models by learning multiple input data models obtained by combining the MAPLE rainfall forecast data, the observed mean rainfall volume data for the area, and the runoff volume data through stored past flood events; a step of forecasting the flood through the flood forecast models by the lead time of the flood forecast and then selecting the flood forecast models forecasting the flood most accurately the lead time; and a step of generating a final flood forecast model by combining the flood forecast models selected by the lead time. The present invention can increase the accuracy of the flood forecast using the MAPLE weather forecast data as short-term rainfall forecast data and all kinds of observation data, and can easily generate and renew the flood forecast models only using the monitoring data and the weather forecast data easily obtained by ANFIS-based flood forecast models as a data-oriented model.