论文部分内容阅读
实现了基于数据挖掘预处理和最小二乘支持向量机的短时用水量预测及监控系统。本系统在原有的基于以太网的自动抄表系统基础上增加了无线自动抄表功能和短时用水量预测功能。该系统具有学习速度快,在小样本情况下具有良好的非线性建模和泛化能力,预测精度高等优点。实践证明此预测及监控系统运行经济可靠。
The short-term water forecasting and monitoring system based on data mining preprocessing and least squares support vector machine is realized. The system adds wireless automatic meter reading function and short-term water consumption prediction function based on the original automatic meter reading system based on Ethernet. The system has the advantages of fast learning speed, good non-linear modeling and generalization ability in small sample cases and high prediction accuracy. Practice has proved that this forecast and monitoring system operating economy and reliable.