基于BP神经网络方法的山东省海洋生态安全评价

Evaluation on marine ecological security in Shandong Province based on BP neural network

基于2002—2012年山东省海洋生态安全的相关数据，从生态安全、经济安全和社会安全3方面选取26个指标构建海洋生态安全评价指标体系，运用BP神经网络对山东省海洋生态安全发展现状和演变趋势进行评价，并结合评价指数和超载度标准进行分析。结果表明：①BP神经网络通过对历史数据的学习，能够实现海洋生态安全的评价，是一种切实可行的生态安全评价方法。②就各子系统而言，2010～2012年山东省海洋生态安全子系统安全程度呈现上升趋势，经济安全子系统和社会安全子系统的安全程度均呈现下降趋势，综合生态安全程度呈现下降趋势。③总体而言，2010年和2011年山东省海洋生态安全处于较为安全等级，2012年生态安全处于临界安全等级，海洋生态安全出现了恶化的趋势。

Based on the related data of marine ecological security of Shandong Province from 2002 to 2012, this paper constructed a marine ecological security evaluation system for Shandong Province from three aspects which were eco- logical security, economic security and social security. And then used the BP-ANN model to make a evaluation for marine ecological security, and at last made an analysis of marine ecological security according to the evaluation in- dex. The results are shown as follows ： （ 1 ） BP neural network method is a feasible marine ecological security evalua- tion method, which can be well fit the marine ecological security system＇s development evaluation. （2） In terms of each subsystem＇ s security degree, the ecological security system showed a upward trend, while the economic and so- cial security system both showed a downward trend. （3） Overall, the security level of Shandong Province in 2010 and 2011 were both general safety, and it was criticality safety in 2012.