摘要
本文针对水轮机空化特性相对应的空间曲面比较复杂的特点,用径向基函数神经网络建立水轮机空化特性模型,对样本的学习采用Levenberg-Marquardt算法。该模型应用于四川紫坪铺水力发电厂中的水轮机空化特性研究。模拟计算结果与试验结果的比较表明,本文给出的模型能真实地表达水轮机的空化特性,可以在水轮机的计算机辅助选型设计和水轮发电机组的优化运行分析等方面应用。
The radial basic function neural network is adopted to establish the model for describing the cavitation of hydraulic turbine. The Levenberg-Marquardt algorithm is applied to carry out the continuous learning of the model and the Maflab software is used to develop the computation program. The simulation result of nonlinear relationship among critical cavitation index, unit rotation and unit discharge for describing the cavitation characteristics of the turbine in Zipingpu Hydropower Station, Sichuan Province is in good agreement with the model test result. It shows that the proposed model is feasible.
出处
《水利学报》
EI
CSCD
北大核心
2006年第7期893-897,共5页
Journal of Hydraulic Engineering
基金
河北省水利科研与推广计划项目(2005-38)
