基于RBF神经网络模型的资源型城市生态安全预警——以榆林市为例

Ecological Security Warning for Resource-Based Cities Based on RBF Neural Network Model: Taking Yulin City as an Example

运用SDI(状态—胁迫—免疫)模型构建资源型城市生态安全预警指标体系,基于RBF神经网络模型及警情评判标准对榆林市2010—2016年的生态安全状况进行分析,并对榆林市2017—2021年的生态安全演化趋势进行动态预测。研究结果表明:(1)RBF神经网络模型能够有效地用于资源型城市生态安全预警分析,拟合精度较高;(2)2010—2016年,榆林市整体生态安全警情指数不断上升,但是仍处于临界安全状态;(3)根据未来趋势预测结果,到2020年,榆林市生态安全"状态"子系统警情指数将有大幅提升,"胁迫"子系统生态安全警情指数呈略微下降趋势,"免疫"子系统呈波动趋势。由此表明,快速提升榆林市生态安全状况,首先要改善胁迫子系统的状况,其次是加强免疫子系统的免疫能力,据此提出了针对榆林市提升生态安全状况的四条响应措施。

The SDI(state-stress-immunization)model was used to construct a resource-based urban ecological security early warning indicator system.Based on the RBF neural network model and the police evaluation criteria,the ecological security status of Yulin City from 2010 to 2016 was analyzed,and the ecological security evolution trend of Yulin City from2017 to 2021 is dynamically predicted.The results show that:(1)RBF neural network model can be effectively used for resource-based urban ecological security early warning analysis,and the fitting accuracy is high;(2)In 2010-2016,the overall ecological security early warning index of Yulin City continues to rise,but still In the critical security state;(3)According to the forecast of future trends,by 2020,the warning index of the'state'subsystem of the ecological security of Yulin City has been greatly improved,and the ecological security early warning index of the'stress'subsystem has a slight downward trend,'immune'subsystems are fluctuating.This indicates that the rapid improvement of the ecological security status of Yulin City must first improve the status of the stress subsystem,followed by strengthening the immunity of the immune system.Based on this,four response measures for improving the ecological security status of Yulin City are proposed.