关键词: 森林生态功能, 空间格局, BP神经网络, 均衡度指数, 贵溪市

Abstract: BP Neural Network Model of forest ecological function in sublot scale was built to analyze the spatial pattern of forest ecological function,a Forest Ecological Function Model in regional scale was built to analyze the overall forest ecological function of the study area and a spatial equilibrium index model of forest ecological function was built to analyze the equilibrium distribution of forest ecological function in the study area. The results show that the BP neural network structure was “8-5-1”, and the network performance could meet the research requirements. The forest ecological function of the study area was divided into three grades of good, medium and poor, which respectively occupied 3.6%, 88.4% and 8.0% of the total forest land area. The score of forest ecological function in the regional scale was 0.609 7, and the grade of overall forest ecological function in study area was poor. The spatial equilibrium index of forest ecological function in the study area was 0.863 8, and the spatial distribution of forest ecological function was relatively balanced. The sublots with middle grade forest ecological function accounted for a large portion in study area, which explaining that the protection of national ecological forest had achieved remarkable results. Forest ecological functions were more spatially distributed, but overall forest ecological functions are poor.

Key words: forest ecological function, spatial pattern, BP neural network, spatial equilibrium index, Guixi city