| 遗传算法优化的BP神经网络城市不透水层百分比估算 |
| |
| 引用本文： | 骆成凤.遗传算法优化的BP神经网络城市不透水层百分比估算[J].测绘科学,2011,36(1):48-50. |
| |
| 作者姓名： | 骆成凤 |
| |
| 作者单位： | 中国测绘科学研究院对地观测技术国家测绘局重点实验室; |
| |
| 基金项目： | 对地观测技术国家测绘局重点实验室开放基金项目资助 |
| |
| 摘 要： | 本研究利用基于遗传算法优化的BP神经网络算法估算城市不透水层百分比。首先,将像元中各端元组分与BP神经网络的节点相对应进行BP网络建模,遗传算子建模;其次,对样本进行网络训练,先通过GA算法得到全局近最优网络权重集,然后用梯度下降算法训练网络,直到找到能充分反映特征空间中的数据分布模式的局部最优网络权重集;然后,训练好的网络被应用于整个影像用来估算城市不透水层覆盖百分比。在此基础上,对北京市地表不透水层百分比进行估算。试验结果表明,本研究所用的方法能有效利用中高分辨率遥感影像数据估算城市不透水层百分比。
|
| 关 键 词： | 城市不透水层 百分比估算 BP算法 遗传算法 |
Estimating urban impervious surface percentage with BP neural network bused on genetic algorithm |
| |
| LUO Cheng-feng.Estimating urban impervious surface percentage with BP neural network bused on genetic algorithm[J].Science of Surveying and Mapping,2011,36(1):48-50. |
| |
| Authors: | LUO Cheng-feng |
| |
| Institution: | LUO Cheng-feng(Key Laboratory of Mapping from Space of State Bureau of Surveying and Mapping,Chinese Academy of Surveying and Mapping,Beijing 100039,China) |
| |
| Abstract: | A Back Propagation(BP) network algorithm based Genetic Algorithm(GA) was adopted in the article to estimate urban impervious surface percentage.Firstly,the BP and GA models were build with end-members in the pixel.Secondly,the net was trained with the samples selected,the near best net weights set in the global space could be found by GA,and then BP used grads descends method to search the infinitesimal point in local space until find the best net weights set.Finally the trained net was used to emulate the ... |
| |
| Keywords: | urban impervious surface estimate of percentage BP algorithm genetic algorithm |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
