论文部分内容阅读
针对预报误差与入流大小呈反比趋势,分析了不同级别入流的误差规律,首先通过极小熵确定分布线型,再用极大熵进行参数估计建立了参数之间的关系,与极大似然法在参数估计中具有同样的凹凸点,即精度较一般适线法得以提高。以福建池潭水库为例分析了预报误差基于不同级别入流的分布规律,从而可降低河道或水库的管理风险。
According to the fact that the forecast error is inversely proportional to the inflow size, the error law of inflow at different levels is analyzed. First, the distribution linetype is determined by minimum entropy, and the parameters are estimated by maximum entropy. The relationship between the parameters and maximum likelihood The method has the same bump in the parameter estimation, that is, the accuracy is improved compared with the general fit method. Taking Chondan Reservoir in Fujian Province as an example, the distribution of forecasting errors based on different levels of inflows is analyzed, which can reduce the management risk of rivers or reservoirs.