基于ANFIS-GM的心墙堆石坝变形预测.pdfVIP

水利水电技术第46 卷 2015 年第3 期 基于 ANFIS -GM 的心墙堆石坝变形预测 钟登华，刘吴元，佟大威，刘玉垒，吴斌平，刘肖军 (天津大学水利工程仿真与安全国家重点实验室，天津 300072) 摘要:本文提出采用自适应网络模糊推理系统(adaptive neuro-fuzzy inference system , ANFIS) 优化 灰色理论模型(Grey Model , GM) 的建模方法来研究预测大坝变形。 ANFIS - GM 模型综合考虑了由 于资料不完备、考虑因素不全面而产生的灰色特性和各影响因素与大坝变形之间存在的模糊特性。 该模型相比于 GM 模型不仅考虑了大坝变形的灰色特性，而且还考虑了水位变化速率、填筑速率 与大坝变形的模糊关系。通过心墙堆石坝沉降变形的实例分析，表明该模型比 GM 模型误差更小。 同时，该模型具有处理小样本，自组织、自学习、自适应，模糊推理的综合能力。 关键词:心墙堆石坝;大坝变形;灰色理论;自适应网络模糊推理系统; ANFIS - GM 模型 中图分类号: TV64 l. 43 文献标识码: B 文章编号: 1α泊-0860(2015 )03-∞01-07 ANFIS - GM based-predictioo 00 deformatioo of rockfiIl dam with core w aII ZHONG Denghua , UU Haoyuan , TONG Dawei , UU Yuxi , WU Binping , UU Xiaojun (State Key Laboratory of Hydraulic En喜ineering Simulation and Safety , Tianjin University , Tianjin 3αJ072 ， China) Abstract: It is proposed herein to adopt the modeling method of the adaptive neuro-fuzzy inference system optirnized grey model ( ANFIS - GM) to study and predict dam defonnation. In AN FlS - GM model , the fuzzy characteristics among the grey character- istics due to the incomplete data and incomprehensive consideration of the relevant factors , all the impacting factors and dam de- fonnation 缸e comprehensively considered. Compared with grey model( GM) , this model consider百 not only the grey characteris- tics of dam deformation , but also the fuz勾 relationship in-between the water level changing rate , filling rate and dam deforma- tion , and then bas the performance to process fuzzy system. Through the case analysis on the settlement deformation of rockfill dam with core