基于模糊推理和关联规则分析的河道洪水预报模型

河道洪水预报过程是一项复杂的非线性计算过程.为及时准确地预报下游河道特征值,针对传统模糊推理中存在的规则数过多和模型参数难以确定的问题,提出了一种新的模糊推理河道洪水预报模型.该模型以T-S模糊推理法为基础,通过对历史数据的关联规则分析和建立非线性优化模型确定模糊规则数目和模型参数,最后预报未来出现的流量数值.实例分析表明,基于模糊推理和关联规则分析的河道洪水预报模型易于理解,特别对防洪比较重要的高流量的预报结果较好.

River flow forecasting model based on fuzzy inference and associated rules analysis

The processing of river flow forecasting includes complicated non-linear calculation, how to gain the characteristics of downriver watercourse duly and exactly based on the river flow forecasting model is very important in practice. Aiming at the existing problems on the number of fuzzy rules and parameters in traditional fuzzy inference, a new model based on T-S fuzzy inference engine is proposed to forecast river flow, which confirms rule numbers and model parameters by using associated rules analysis on historical data and non-liner programming method and therefore predicts the future flux value. Through case study, it is testified that the established model based on fuzzy inference and associated rules analysis is easy to understand and implement, especially to excellent precision for flood forecasting.