摘要
针对如何利用生物监测技术提高异常水质识别率的问题,对生物水质评价因子进行了研究,提出了改进的鱼群重心算法。在此算法的基础上引入新的鱼群运动行为特征参数即鱼群离散度、曲率、邻近特征,并对这些特征参数进行预处理,建立水质异常评价因子数据集,最后将评价因子输入支持向量机(SVM)进行水质异常识别。实验结果表明,引入的特征参数用于水质异常评价中,识别效果明显优于其它方法,识别率达到92%以上。
Against how to use biomonitoring technology to improve the recognition rate of abnormal water quality, the evaluation factors of biological water quality is studied and an improved algorithm of fish center is proposed. On the basis of this algorithm, it introduced new behavior characters of fish movement, such as fish dispersion, curvature and vicinity feature, carried out pretreatment for these characteristic parameters, established databases for evaluation factors of abnormal water quality and took the evaluation factors as inputs of support vector machine (SVM) to identify abnormal water quality. The results of experiment show that by introducing characteristic parameters for evaluation of abnormal water quality, its identification effect is better than other methods and the recognition rate is over 92%.
出处
《计量学报》
CSCD
北大核心
2017年第2期175-178,共4页
Acta Metrologica Sinica
基金
国家自然科学青年基金(61601400)
河北省自然科学青年基金(F2012203031)
关键词
计量学
生物监测
鱼群重心算法
评价因子
支持向量机
metrology
biomonitoring
algorithm of fish center
evaluation factors
support vector machine