摘要
综合考虑方案的国民经济评价、技术评价和社会评价等因素,采用高维降维技术投影寻踪分类模型(PPC),利用基于实数编码的加速遗传算法(RAGA)优化其投影方向,将多维数据指标(样本评价指标)转换到低维子空间,根据投影函数值的大小评价出样本的优劣,从而做出决策,最大限度避免了模糊综合评判中权重矩阵取值的人为干扰,取得了满意效果。模型简单、高效、可行。
In this paper, many factors such as the national economy, technology and social evaluations of a project are taking into consideration in the selection of water-saving irrigation schemes. The multi-dimensional dimension-reductlon technology of Projection Pursuit Classification Model （PPC） is utilized to converse the high dimension indexes （Sample Evaluation Index） into low dimension space through using Real number coding Acceleration Genetic Algorithm （RAGA） to optimize their projection direction. The quality of the projection value of the scheme is appraised to make decision. The approach can avoid the artificial disturbance of Fuzzy Corn prehensive Evaluation in the weight value identification to the maximum extent with satisfactory effect. The model is proved to be simple, high-efficient and feasible.
出处
《中国农村水利水电》
北大核心
2006年第12期13-15,共3页
China Rural Water and Hydropower
基金
国家高科技研究发展计划项目(2002AA6Z3263)
关键词
RAGA
PPC
节水灌溉
综合评价
水资源
RAGA
PPC
water-saving irrigation
comprehensive appraisal
water resources
