自适应滤波法在洪水预测中的应用

尝试用时间序列分析中的自适应滤波法对江河灾害性洪水进行预测，认为加权因子个数和自适应常数的确定是关键。以荆江沙市站１９５０～１９９５年的年最高水位资料为研究对象，借助计算机编程与计算，确定了最优加权因子个数和自适应常数，建立了适合于该站未来年最高水位的预测模型。预测结果与误差分析表明，自适应滤波法用于江河洪水预测是可行的。

APPLICATION OF ADAPTIVE FILTERING TO FLOOD FORECASTING

Adaptive filtering in time series analysis is used to predict a stream flood. The key in this process is determination of the adaptive constants and the number o f weighted factors. Taking the highest water levels of the Jingjiang River in Shashi station from 1950 to 1995 as an example, the number of weighted factors and adaptive constants are obtained by programming and calculation and a model suitable to the prediction of the highest water levels in station is developed. The predictive results and error analysis show that the method is feasible for stream flood prediction.