|
|||||
|
|
| 基于LS-SVM的供水管网安全性评价方法研究 | |
| 引用本文： | 王志涛,苏经宇,王威,邢涛.基于LS-SVM的供水管网安全性评价方法研究[J].中国安全科学学报,2010,20(11). |
| 作者姓名： | 王志涛 苏经宇 王威 邢涛 |
| 作者单位： | 1. 北京工业大学,北京城市与工程安全减灾中心,北京,100124;北京工业大学,建筑与城市规划学院,北京,100124 2. 北京工业大学,建筑与城市规划学院,北京,100124 3. 北京市城市系统工程研究中心,公共安全重点实验室,北京,100089 |
| 基金项目： | 国家"十一五"科技支撑计划项目，北京城市系统工程研究中心综合风险评估实验室建设项目(PXM2009 178215 0893471216).北京工业大学博士科研启动基金资助 |
| 摘 要： | 为评价城市供水管网的安全性,保障其正常运行,笔者基于多元分类最小二乘支持向量机(LS-SVM)的方法,在对城市供水管网安全运行影响因素总结与分析的基础上,构建供水管网安全性评价的指标因素集与评价模型,通过对有限的经验数据的学习,建立供水管网安全性与其影响因素之间的非线性关系。运用该模型进行实例仿真模拟,通过与实际安全等级及BP神经网络模型预测安全等级之间的对比表明:基于LS-SVM的供水管网安全性评价方法具有较高的精度,正确分类率可以达到83.33%。 |
| 关 键 词： | 供水管网 供水安全 安全评价 最小二乘支持向量机(LS-SVM) 人工神经网络(ANN) |
Study on Safety Evaluation Method for Water Distribution Networks Based on LS-SVM |
|
| WANG Zhi-tao,SU Jing-yu,WANG Wei,XING Tao.Study on Safety Evaluation Method for Water Distribution Networks Based on LS-SVM[J].China Safety Science Journal,2010,20(11). | |
| Authors: | WANG Zhi-tao SU Jing-yu WANG Wei XING Tao |
| Abstract: | To evaluate the security of urban water distribution network and ensure its normal operation,a model based on LS-SVM was proposed.On the basis of the summary and analysis of influential factors for urban water distribution network security,a set of indexes used in the above evaluation model was constructed.The nonlinear mapping between the water distribution networks security classification and its conditions was constructed by learning from the finite samples.A water distribution network example was simulated using this model;and the BP ANN model was also used to simulate the same example.A comparison among the actual security level,the one by LS-SVM model and the one by BP ANN model was made.It is concluded that the result gained by LS-SVM model is higher in accuracy,the correct classification rate may reach 83.33%. |
| Keywords: | water distribution networks water-supply safety safety evaluation least squares support vector machines(LS-SVM) artificial neural networks(ANN) |
| 本文献已被 CNKI 万方数据 等数据库收录！ | |