滴灌滴头水力性能预测模型  

Prediction model for hydraulic performance of trickle irrigation emitters

作　　者：刘春景[1,2] 郑加强[2] 周宏平[2] 胡天翔[2] 

机构地区：[1]蚌埠学院机械电子工程系,安徽蚌埠233030 [2]南京林业大学机械电子工程学院,江苏南京210037

出　　处：《水利学报》2010年第11期1353-1359,共7页Journal of Hydraulic Engineering

基　　金：国家"十一五"科技支撑计划课题(2006BAD08A1203);安徽省教育厅科研项目(2007jq1126)

摘　　要：将基于改进粒子群优化最小二乘支持向量机的预测模型引入滴头水力性能预测领域,对所建立的滴头水力性能预测模型分别从理论和试验两方面进行验证。结果表明,在相同样本条件下,该模型构造速度比标准LS-SVM方法快近1个数量级,模型预测误差约为标准LS-SVM方法的50%,预测精度比常规BP模型高1个数量级。本文所建立的预测模型能较好地描述流道宽度、流道半长、流道转角和流道单元数等4个因素对滴头水力性能的影响。A novel prediction model for hydraulic performance of trickle irrigation emitter based on modified PSO least square support vector machine is proposed and verified by experiment. The research results show that the construction speed of this modified PSO LS-SVM model is lO times faster than that of common LS-SVM model,while the prediction errors were reduced about 50%. Moreover, compared with BP model, the prediction accuracy is about 10 times better than that of the former. The proposed prediction model can explain the influence of width, length, and angle and unit number of emitter channel to hydraulic performance.

关 键 词：滴灌滴头 水力性能 预测模型 改进粒子群优化最小二乘支持向量机 

分 类 号：S275.6[农业科学—农业水土工程]