摘要
介绍BP神经网络评价原理、方法,并以洞庭湖湿地生态系统健康评价为例,对评价指标体系构建、湿地生态系统健康度设定、BP学习训练、BP神经网络的可行性检验进行了论述.评价结果,2001~2003年洞庭湖湿地生态系统的健康度分别为0.64、0.62和0.58,健康程度从亚健康向较不健康转变,认为必须及时采取合理的保护及湿地资源开发利用措施以逆转恶化趋势.
The paper introduces the principle and method for BP neural network assessment, taking healthy assessment of wetland ecosystem of Dongting Lake as an example, and elaborates assessment indi- cator system, healthy assumption, BP learning and training, the feasibility of BP neural network. The re- sult of the assessment reveals that the healthy degree of the lake＇ s ecosystem is 0.54,0.62, and 0.58 from 2001 to 2003 in order, tending to transform from sub-healthy to unhealthy. It concerns that it is necessary to take effective measures to protect and develop wetland resources in order to change the trend for worsening.
出处
《林业调查规划》
2009年第5期47-50,共4页
Forest Inventory and Planning
基金
湖南省教育厅重点科研项目<湖南省湿地资源分类与评价>
关键词
BP神经网络
湿地生态系统
健康评价
洞庭湖
BP neural network
wetland ecosystem
healthy assessment
Dongting Lake
