摘要
在为实现对物流企业汽车油耗数据的全面实时监控而开发的软件系统中,基于.Net技术实现了对于汽车异常油耗数据的降噪、自动识别、判断、捕捉与显示功能。在此过程中,首先对汽车油耗数据的特点进行分析,选取决策树分析法对数据进行分类,并在结果中区分不同类型的异常釉耗数据。软件使用C#代码编写友好的可视化界面,并以表格和图形两种方式提交查询结果,便于用户操作。
In the software system that＇s programmed to monitor the usage of vehicles＇ fuel consumption in logistics enterprises, a series of functions are developed based on . Net, which contains how to classify, estimate, capture and display the outliers of the motor vehicle＇s fuel consumption. Firstly, the features of the motor vehicle＇s fuel consumption are analyzed. Then the decision tree algorithm is applied to realize the classification of all the data and identify the different types of the outliers in the classified data. The software system uses C# language to program a friendly and visual interface, and submit the result to users in easy-operating form of table and picture.
出处
《机电一体化》
2010年第11期68-71,共4页
Mechatronics
关键词
异常数据
决策树
自动捕捉
降噪
outlier decision tree automatic capture noisy data
