摘要
分析了畸变波形下感应式电能表和全电子式电能表的计量误差,引入BP神经网络建立电能表计量误差数学模型。用VC++.net开发了谐波电能计量误差分析软件,并应用多线程技术,把神经网络模块和计算模块等放在主线程中实现,而把数据采集模块放在后台工作线程中。采用加汉宁窗和插值改进的傅里叶变换,分析各次谐波电流和电压的幅值、相位。该软件可实现手工输入数据计算,也可配合数据采集系统实时计算出各次谐波对电能计量装置的计量误差和总计量误差。该软件已投入实际运行且情况良好。
The energy metering errors of induction meters and electronic meters distorted waveforms are analyzed. A mathematical model of energy metering error is built up with BP neural network and an analytic software package for harmonic metering error is developed with VC++.Net. Multi-thread technology is adopted,which puts the neural network module and calculation module into the main thread,and the data acquisition module into background thread. The improved FFT algorithm with windows and interpolation is used to calculate the amplitudes and phases of harmonic currents and voltages. Manual data entry is permitted,and cooperated with data acquisition system the metering error of any order harmonic ,and the total metering error as well, can be calculated. The designed software is applied in practice and runs well.
出处
《电力自动化设备》
EI
CSCD
北大核心
2006年第1期76-79,共4页
Electric Power Automation Equipment