Song Lin

An Undersampling Digital Spectral Wattmeter with Power Quality Monitoring

May 22, 1998
10:00 AM
406 Egan

Abstract

In the emerging new market for electric power, power deliverers will sell power quality as well as power quantity. It thus becomes essential to be able to measure power quality i.e., harmonic content and temporal transients, in addition to the energy delivered. Conventional analog watt-hour meters, which are based upon the principle of mechanically averaging instantaneous power, have no means of meeting such requirements. We present a Digital Spectral Watt-hour Meter that has the following features: (i) accurate measurement of power consumption, (ii) measurement of harmonic content (iii) monitoring of transients, sags, brown-outs, dropouts, (iv) digital output, which can be used for summary reports and remote data acquisition, (v) low computational complexity, and potentially low cost in mass production.

Our design is based upon the combination of monitoring and digitizing voltage and current signals, a design of a Motorola 68000-based embedded processor system and an efficient signal processing algorithm. The main innovation in this algorithm is its ability to accurately determine the harmonic content of a periodic waveform, while using a sampling rate well below the corresponding Nyquist rate.

Thesis Committee:
Prof. H. Lev-Ari (Advisor)
Prof. B. Shafai
Prof. A.M. Stankovic