摘要
传统水质监测模型的识别模式,仅依靠单一的分类方法确认水质信息,训练与学习能力较差,直接影响模型监测精度,设计了基于数据挖掘的城市人居环境河流水质变化监测模型。预先计算影响水质变化的参考指标,利用数据挖掘方法,优化对水质变化的识别模式;结合时间序列,建立具有时间性质的水质变化监测模型。实验结果表明:此次设计的监测模型学习与训练能力更强,其相对误差绝对值平均结果为7.13%,泛化误差大幅度下降,有更好的监测性能,可以较好地完成城市人居环境河流水质变化监测。
The recognition mode of traditional water quality monitoring model only relies on a single classification method to confirm the water quality information,so the training and learning ability is poor,which directly affects the monitoring accuracy of the model.Therefore,this study designed a monitoring model of river water quality change in urban living environment based on data mining.Firstly,the reference indexes affecting water quality change were calculated in advance,and the recognition mode of water quality change was optimized by using data mining method;Combined with time series,a monitoring model of water quality change with time property was established.The experimental results showed that the designed monitoring model has stronger learning and training ability,the average absolute value of relative error is 7.13%,the generalization error is greatly reduced,and has better monitoring performance,which can better complete the monitoring of river water quality change in urban living environment.
作者
史利涛
SHI Li-tao(Linking Sub-Bureau of Luohe Ecological Eiwironment Bureau,Luohe,Henan 462600,China)
出处
《四川环境》
2022年第4期219-224,共6页
Sichuan Environment
关键词
数据挖掘
城市人居环境
河流水质
监测模型
Data mining
urban living environment
river water quality
monitoring model