基于OCR技术的电网运行异常数据修正模型

Abnormal data modified model of power grid operation based on OCR technology

针对电网运行数据结构复杂、容易产生数据异常的问题,设计基于OCR技术的电网运行异常数据修正模型。使用基于OCR技术的电网运行数据采集方法获取电网运行数据,结合径向基神经网络采集数据负荷曲线的二次聚类,提取日负荷曲线特征,在其相似性和平滑性的基础上,运用基于综合云的电网运行异常数据修正模型修正异常数据。实验结果表明:该方法的异常数据修正值与实际值基本一致,修正能力较强,修正效率较高。

Since the problem that the power grid operation data structure is complex and prone to abnormal data,a power grid operation abnormal data model based on OCR technology is designed.The power grid operation data acquisition method based on OCR technology is used to obtain the power grid operation data.Combined with the secondary clustering of data load curve collected by radial basis function neural network,the characteristics of daily load curve are extracted.On the basis of its similarity and smoothness,the power grid operation abnormal data model based on comprehensive cloud is used to correct the abnormal data.The experiment results show that the abnormal data correction value of this method is basically consistent with the actual value,this method has strong correction ability and high efficiency.