摘要
以地下水各聚类指标水质级别国家标准值为评价标准,应用基于高维降维技术的投影寻踪模型,并采用改进的免疫蛙跳算法对投影寻踪模型投影方向进行优化求解,将高维数据指标转换到低维子空间,根据计算得出投影函数值的大小对样本水质进行评价。结果表明:免疫蛙跳算法收敛速度、搜索精度优于混合蛙跳算法;基于免疫蛙跳算法优化的投影寻踪模型能够很好的对水环境质量进行评价,可用于其他环境的综合评价。
Based on the national standard for groundwater quality,the multi-dimensional dimension-reduction technology of projection pursuit classification model(PPC) is adopted.Improved immune shuffled frog leaping algorithm(ISFLA) is used to optimize their projection direction.The high dimension index is conversed into low dimension.The projection function value is used to appraise water quality.The results show that convergence speed and searching precision of immune shuffled frog leaping algorithm are better than those of shuffled frog leaping algorithm.The model can assess water quality appropriately.
出处
《中国农村水利水电》
北大核心
2010年第5期5-7,共3页
China Rural Water and Hydropower
基金
国家科技支撑项目(2007BAD88B01-06)
关键词
水质评价
混合蛙跳算法
免疫算法
投影寻踪
water quality
shuffled frog leaping algorithm
immune algorithm
projection pursuit