论文部分内容阅读
大气物理参数是影响漓江的水位变化的主要因素,对2000～2006年连续各月桂林市的大气温度、大气相对湿度、降水量、日照时数等大气物理参数和桂林水文站漓江水位数据,使用相关分析法分析大气物理参数对漓江水位变化的影响程度,并应用径向基人工神经网络由大气物理参数对漓江水位变化进行预测。结果表明,造成漓江水位变化的最重要大气物理参数是降水量,次之为大气温度;正常气象情况下,根据降水量和大气温度利用径向基神经网络可较为准确地预测漓江年最低水位和年最高水位,预测最低水位的误差小于5%,预测最高水位的误差略大于5%,预测最低水位比最高水位更准确,可为漓江流域可能发生的旱情、洪水等问题提供科学的决策数据。
The atmospheric physical parameters are the main factors affecting the water level change in Lijiang River. Using the atmospheric physical parameters such as atmospheric temperature, relative humidity, precipitation, sunshine duration and other data of Guilin Lijiang River water level in successive successive months from 2000 to 2006, Correlation analysis was used to analyze the influence of atmospheric physical parameters on the water level change of Lijiang River. Radial basis artificial neural network was used to predict the water level change of Lijiang River from atmospheric physical parameters. The results show that the most important atmospheric physical parameter causing the Lijiang River water level change is precipitation and the second is the atmospheric temperature. Under normal weather conditions, the radial basis function neural network can accurately predict the minimum water level The annual maximum water level is less than 5% of the predicted minimum water level error, and the error of the maximum water level forecast is slightly more than 5%. The predicted minimum water level is more accurate than the maximum water level, which can provide scientific decision-making data for possible drought and flood in the Lijiang River Basin.