Application Of Neural Network In Ecological Security Evaluation Of Urban Landscape

Cities are both natural creations of human civilization and systems in which humans cohabit with other organisms.Because the urban landscape ecosystem is the most fundamental carrier of urban planning,the success of urban landscape ecological safety management has a direct impact on the fitness of the urban planning system for the survival of various people and organisms.Changes in China’s urban landscape ecology are visible as a result of the world’s rapid population expansion and expanding urbanization,and there are still significant opportunities and difficulties for urban landscape ecological safety issues.The protection of urban landscape ecological security is the top goal of environmental governance since it is a basic guarantee for China’s homeland security,economic and social sustainable growth.However,scientific study on urban landscape ecological challenges in China is still in its early stages,with no system in basic theory and no systematic and comprehensive planning system in place.The primary cause of this scenario is that people’s understanding of environmental issues is insufficiently broad and profound.It is vital to address the relationship between high-quality social development and environmental conservation,particularly to acknowledge the contradiction between urban economic and social development and environmental protection.This manuscript aims to investigate the application of neural network models in urban landscape ecological security evaluation,introduce and apply the entropy-PCA method and the BP neural network method in urban landscape ecological security evaluation,and use Python programming language to conduct computer modeling for the algorithm used in this study,in order to quantitatively analyze the urban landscape ecological security level.Finally,in Zhengzhou,the application of neural networks in urban landscape ecological security rating is empirically examined.The following are the key conclusions of this manuscript.(1)Based on the framework of " Drivers-Pressures-State-Impact-Response "(DPSIR),we construct a suitable index system for urban landscape ecological security evaluation.Based on the principle of index selection and the characteristics of urban landscape ecosystem,this paper starts from five criteria: driving force,pressure,state,impact and response,and divides them into 12 elements: economic development,social development,population pressure,resource pressure,environmental pressure,ecological state,climate state,environmental impact,economic impact,human impact,measure response and economic response.A total of 27 indicators were finally selected to refine the measurement parameters of urban landscape ecological security evaluation,laying the foundation for the subsequent evaluation of urban landscape ecological security.(2)Introducing the entropy-PCA method in the BP neural network,the rigor of the landscape ecological safety evaluation results is modified.The effectiveness and accuracy of BP neural network applied to urban landscape ecological safety evaluation were verified by conducting example application,using the example data and the results obtained from the example operation as the training data set of BP neural network,and comparing with the entropy-PCA method.(3)The application of BP neural network model in urban landscape ecological security evaluation was verified empirically.Based on the statistical data of Zhengzhou City from2000-2019,the DPSIR model for landscape ecological security evaluation of Zhengzhou City was established,and the algorithm modeling was carried out using the Python programming language,and the landscape ecological security index was calculated for each study year based on the entropy weight method coupled with the PAC principal component analysis method,and the BP neural network model was used to optimized and rated for analysis.(4)In the case study,the entropy-PCA method was used to select the 27 most important and comprehensive evaluation indicators.At the criterion level,pressure indicators,state indicators and response indicators all have a large impact on the landscape of Zhengzhou.The driver indicator factor had the greatest impact,followed by the impact and response indicator factors,and the state indicator factor had the least impact,indicating that the effectiveness of Zhengzhou in mitigating the effects of social and economic development is a focus of future efforts to maintain ecological security in the study area.(5)Through the case study,we found that the landscape ecological security trend of Zhengzhou City from 2000 to 2019: the landscape ecological security level of Zhengzhou City has experienced three types of levels,and the comprehensive evaluation level of entropyPCA method and the comprehensive evaluation level of BP neural network have experienced level Ⅰ,level Ⅱ and level Ⅲ.The entropy-PCA method comprehensive evaluation grade of landscape ecological security in Zhengzhou City was at level Ⅱ for 8 consecutive years from2000 to 2007;it dropped to level Ⅲ in 2008;it rebounded slightly to level Ⅱ in 2009 and 2010;it dropped to level Ⅲ again in 2011 and 2012;it continued to fluctuate and increase from 2013 to 2018.-The comprehensive evaluation index fluctuates and decreases to Ⅲ in 2019.Based on the comprehensive evaluation index obtained by the entropy-PCA method,the BP neural network comprehensive evaluation index is calculated and divided to obtain the BP neural network comprehensive evaluation grade.The division of the basic level of landscape ecological security in Zhengzhou from 2000 to 2018 is consistent with the comprehensive evaluation level of the entropy-PCA method,which has experienced three level shifts,fromⅡ level down to Ⅲ level and then fluctuating up to Ⅰ level;different from the comprehensive evaluation level obtained by the entropy-PCA method,Zhengzhou 2019 BP neural network comprehensive evaluation index is 0.633 5,and the landscape ecological security level is classified as level Ⅱ.(6)The entropy-PCA method and BP neural network are used to calculate the comprehensive evaluation indexes of landscape ecological security in Zhengzhou,and the results of comparative analysis show that the results obtained by the two methods are basically consistent,and the two methods are mutually verified and integrated,which makes the evaluation results of urban landscape ecological security more scientific and rigorous.(7)To propose sub-regional regulation and control and overall landscape ecosystem security level countermeasures to promote the landscape ecological security posture of Zhengzhou city to change to excellent.Accelerate the transformation of agricultural industry in eastern Zhengzhou,promote the construction of ecological barrier in western Zhengzhou,improve the traffic hub pattern in southern Zhengzhou,guarantee the high-quality development along the yellow in northern Zhengzhou,control the scale of construction in central Zhengzhou,and achieve the effect of optimizing urban landscape ecological security as a whole.This manuscript employs a mix of qualitative and quantitative methods to investigate the application of an urban landscape ecological security evaluation model,using Zhengzhou as an example study,evaluating its landscape ecological security level,analyzing the main influencing factors of Zhengzhou city landscape ecological security,and proposing measures to regulate Zhengzhou city landscape ecological security on this basis,providing a reference for urban landscape ecological security.This study provides a scientific foundation and direction for the development of ecological civilization in the Central Plains city cluster.