Calibration of RMS based on hardware compensation method and Least Squares Support Vector Machine
Download as PDF
DOI: 10.25236/icscbd.2018.009
Author(s)
Yan Huang, Anquan Deng, Tianyu Zhou, Wu Zhu
Corresponding Author
Wu Zhu
Abstract
When the RMS measurement is performed on a wideband AC signal using a log-antilog operation circuit, a large nonlinear error is generated by the band limitation of the amplifier in the converter. Therefore, a new method for calibrating nonlinear errors using hardware calibration and software alignment is proposed. First, the hardware circuit with the voltage-frequency conversion chip VFC32 as the core is used for nonlinear calibration. The calibrated signal is then fitted using a least squares support vector machine. The test results show that the calibration accuracy can be achieved 0.014% within the frequency range of 50~20 ×103Hz.
Keywords
Rms Converter, Nonlinear Error, Error Calibration, Hardware Compensation Method, Least Squares Vector Machine Algorithm