热带地理Issue(4)：373-380,406,9.
基于改进遗传算法投影寻踪的大城市郊区耕地安全综合评价
Comprehensive Evaluation of Arable Land Security in Metropolitan Suburbs Using Projection Pursuit Model Based on Improved Genetic Algorithms
摘要
Abstract
Aiming at the problems of poor objectivity of multidimensional index weight assignment for evaluation of arable land security, the authors proposed and applied projection pursuit model (pp) optimized by a real coded accelerating genetic algorithm (RAGA) to evaluate the arable land security in the rapid urbanization suburbs of Metropolitan cities. In the paper Panyu district and Conghua city in Guangzhou are taken as examples, and 32 indicators are selected to build four security sub-systems, namely, the quantity security, structural security, quality security, ecological security, and economic and social security. The results indicate that: (1) Using RAGA-PP model can achieve satisfactory results in the objective evaluation of multi-target and multi-factor arable land security.(2) The weights of arable land security impact index between outer suburbs and peri-urban area in Guangzhou city are very different. (3) Due to the difference of development orientation and development strategy, the urbanization rates of the areas are different, the arable land security level of Panyu is lower than that of Conghua, as Panyu is in peri-urban area while Conghua in outer suburbs. (4) From the sub-goals, Panyu’s arable land has a higher level of economic security, but it’s quantity security, quality and ecological security are lower than those in Conghua.关键词
耕地安全/投影寻踪法/实码加速遗传算法/大城市郊区/番禺/从化Key words
arable land security/projection pursuit method/real coded accelerating genetic algorithm/Metropolitan Suburbs分类
信息技术与安全科学引用本文复制引用
王芳,冯艳芬,卓莉,周涛..基于改进遗传算法投影寻踪的大城市郊区耕地安全综合评价[J].热带地理,2013,(4):373-380,406,9.基金项目
国家自然科学基金资助（41001048，40801034）；广州市哲学社会科学“十一五”规划资助项目 （）
