期刊名称：IAENG International Journal of Computer Science
印刷版ISSN：1819-656X
电子版ISSN：1819-9224
出版年度：2020
卷号：47
期号：1
语种：English
出版社：IAENG - International Association of Engineers
摘要：The land ecological security evaluation should be a research emphasis for its key role in the sustainable development of a region. In this research, according to the PSR framework-based land status of Yuxi City, an evaluation index system has been built up to reveal the changes of the land ecological security and analyze the causes of the variations from 2001 to 2015 in Yuxi City. This system is composed of three layers--direction layer, criterion layer which further consists of pressure, status and response, and index layer with 20 indices covering the various aspects of land use, society, economy and environment. In this paper, the genetic algorithm (GA) is introduced to improve the BP neural network, with advantages in solving the problems of slow convergence and getting into local minimum easily when the BP neural network was applied alone in land ecological security evaluation. A GA-BP neural network is then established to evaluate the land ecological security from 2001 to 2015. Comparisons between the BP neural network and the GA-BP neural network are drawn in their performances and errors and the assessment results of both are further separately compared with the target results of the comprehensive index method. The results show that: (1) The land ecological security index increases steadily from 0.3696 to 0.6020 and the security grade ascends from risky (IV) to safe (II) from 2001 to 2015 in Yuxi City; (2) Compared to the traditional BP neural network, the GA-BP neural network has less errors in training and predicting, and it is faster in convergence and higher accuracy in assessing results. Therefore, the GA-BP neural network model is not only able to function as well as the BP neural network in land ecological evaluation and prediction, it can obtain more accurate results and has faster convergence ability as well.
关键词：BP neural network;genetic algorithm;land ecological security evaluation;Yuxi City
