Water Science & Technology Vol 66 No 2 pp 239–246 © IWA Publishing 2012 doi:10.2166/wst.2012.161

Application of hydrologic forecast model

Xu Hua, Xue Hengxin and Chen Zhiguo

In order to overcome the shortcoming of the solution may be trapped into the local minimization in the traditional TSK (Takagi-Sugeno-Kang) fuzzy inference training, this paper attempts to consider the TSK fuzzy system modeling approach based on the visual system principle and the Weber law. This approach not only utilizes the strong capability of identifying objects of human eyes, but also considers the distribution structure of the training data set in parameter regulation. In order to overcome the shortcoming of it adopting the gradient learning algorithm with slow convergence rate, a novel visual TSK fuzzy system model based on evolutional learning is proposed by introducing the particle swarm optimization algorithm. The main advantage of this method lies in its very good optimization, very strong noise immunity and very good interpretability. The new method is applied to long-term hydrological forecasting examples. The simulation results show that the method is feasibile and effective, the new method not only inherits the advantages of traditional visual TSK fuzzy models but also has the better global convergence and accuracy than the traditional model.

Keywords: hydrologic forecast; PSO; TSK fuzzy system; visual principle; Weber law

Full article (PDF Format)