The Spatio-temporal Simulation Of Regional Ecological Security Based On Neural Network

Ecological security,an important cornerstone of national security,is a not damaged or threatend state of the ecological environment for a country's survival and development.With the growing social economy,some regions have been facing up with serious issues,such as the competitive land use between economic and ecological construction,the weakening ecological resources carrying capacity and the aggravation of ecological damage and pollution,strongly affecting and restricting their regional ecological security and sustainable development.Therefore,it is of great theoretical and practical significance for the protection and construction of regional ecological security under the new normal to start from the regional scale,analyzing the state of ecosystem security,predicting and simulating the development trend of regional ecological security scientifically,and then putting forward the corresponding reasonable solutions.This paper takes Shaanxi Province as the study area with various natural conditions,complex social and economic situations.Firstly,the evaluation index system of ecological security was constructed based on the framework of "Pressure-State-Response".Secondly,entropy weight method was used to determine the index weight,and the parameters such as regional ecological security pressure index,ecological security state index,ecological security response index,and ecological security comprehensive index were calculated by weighted synthesis method,then their temporal and spatial differences were analyzed.Finally,the Back Propagation Neural Network model was constructed with Maxout as the activation function.Based on the simulation and verification,the ecological security pressure index,ecological security state index,ecological security response index and ecological security comprehensive index of Shaanxi Province from 2016 to 2021 were predicted,and their spatiotemporal differences were analyzed.The conclusions are as follows:(1)Based on the framework of "Pressure-State-Response",the evaluation index system of regional ecological security suitable for Shaanxi Province isconstructed and divided into target level,criterion level and indicator level from the top to the bottom.The criterion level is divided into three subsystems: pressure,state and response.The indicator level is a specific and measurable indicator that represents the regional ecological security status.Combined with the actual situation of Shaanxi Province,20 indicators that can reflect the state of three criterion levels are selected.Besides,the data is processed through the index standardization and entropy weight method,which lays the foundation for the subsequent regional ecological security evaluation of Shaanxi Province.(2)From 2005 to 2015,the overall ecological security of Shaanxi Province has improved significantly.From the low safety level in 2005 to the high safety level in 2015,the spatial distribution pattern has changed from "risk and comparative safety coexist in the northern,severe in the central,and risk in the southern" to "critical safety,comparative safety in the northern and southern regions,critical safety in the central".Among them,the spatial distribution of pressure level in Shaanxi Province has changed from the pattern of "comparative safety in the northern and southern,and severe in the central" to the pattern of "critical,comparative safety in the northern,critical safety in the central,and relatively safety in the southern";the spatial distribution of state level has changed from the pattern of overall severe to the pattern of "high safety level of critical safety and above in the northern and southern,and low safety level of critical safety and below in the central";the spatial distribution of response level changes from the pattern of overall severe to the pattern of "risk and comparative safety coexist in the northern,comparative safety in the central,and critical safety in the southern".(3)The simulation accuracy of Maxout-BP neural network model is high.Based on the new activation function Maxout,an improved BP neural network is constructed.On this basis,the grid data of ecological security pressure index,state index,response index and comprehensive index of Shaanxi Province in 2015 are simulated and compared,with the accuracy of 85.93%,80.55%,80.38% and 83.98%,respectively.The accuracy of verification is above 80%,establishing a foundation for the prediction of ecological security in Shaanxi Province.(4)During 2016 to 2021,the overall ecological security situation of Shaanxi Province has been basically stable,with the spatial distribution pattern turning from "critical safety,comparative safety in the northern and southern regions,and critical safety in the central" to"critical safety,comparative safety in the northern and southern regions,and comparative safety in the central".Among them,the spatial distribution pattern of pressure level in Shaanxi Province has changed from "comparative safety in the northern,critical safety in the central,and relatively safety in the southern" to "low safety level of critical safety and below in the northern and central,and comparative safety in the southern";the spatial distribution of state level has changed from the pattern of "high safety level of critical safety and above in the northern and southern,and low safety level of critical safety and below in the central" to the pattern of "relatively safety in the northern,critical safety in the central,and comparative safety in the southern";the spatial distribution of response level has been changed from the pattern of "risk and comparative security coexist in the northern,comparative security in the central,and critical security in the southern" to the pattern of "relatively safety in the northern and southern,comparative safety in the central".