论文部分内容阅读
介绍了建立最优模糊划分自激励门限自回归模型的观点、理论和方法。该建模方法克服了传统方法确定门限自回归模型门限值的困难，为用线性化方法对非线性时间序列的建模和预报提供了一种新的数学方法，具有广泛的实用性。建立了６种数据处理方式下的泾河、北洛河、渭河及大通河主要水文站的候、旬、月枯水径流预报模型。模型评定与检验显示，将时段平均流量序列转换为消除年周期的模比系数序列建立的模型优于其他类型的模型，可作为作业预报模型。
The viewpoints, theories and methods of establishing the optimal fuzzy self-excitation threshold autoregressive model are introduced. The modeling method overcomes the difficulty of determining the threshold value of the threshold autoregressive model by the traditional method, and provides a new mathematical method for modeling and forecasting nonlinear time series by the linearization method, which has a wide range of practicability. Six kinds of data processing methods were established for the forecasting models of the runoff, litter and monthly runoff of Jinghe, Beiluohe, Weihe and Datonghe hydrological stations. The model evaluation and test showed that the model established by converting the time series average flow rate series into the model cycle with the elimination of annual cycles is superior to other types of models and can be used as the job forecasting model.