Energy Metering Algorithms And Their SoC Implementations

This thesis takes designing a high accuracy energy metering system on chip (SoC) under harmonic distortions as the goal, focusing on research of energy metering algorithms suitable for application-specific integrated circuit (hereinafter referred to as ASIC) implementations. The achievements include:(1) Modeled the energy metering system to detect the error sources and discussed linear error and jitter error on impact of energy metering accuracy separately. (2) Analyzed the advangtages and disadvangtages of most commonly used harmonic analysis algorithms-fast Fourier transform (hereinafter referred to as FFT) and wavelet packet decomposition for the application of harmonic energy metering system, then proposed a dedicated digital filter based harmonic energy metering algorithm with relatively lower hardware cost.(3) Used a number of specified digital signal processors (hereinafter referred to as DSP) to implement the dedicated digital filter based harmonic energy metering algorithm in an SoC, which also integrates eight over sampled analog to digital convertes (hereinafter referred to as OSADC), a microprocessor control unit and its peripheral interfaces. Testing results of the energy meter based on the above SoC indicate that its accuracy can meet the requirement of 0.2S national stardard rate. (4) Proposed an all digital synchronous sampling algorithm. The algorithm features in dynamically adjusting the decimation ratio of the decimation filter of the OSADC by the fluctuation of electrical power's fundamental frequency. With this algortihm the metering outputs from FFT can reach high accuracy without window or interpolation calculation. (5) Presented a low cost and high steady-state accuracy frequency estimation algorithm with moderate response speed to aid the circuit implementation of the all digital synchronous sampling algorithm. Both the dedicated digital filter based algorithm and the all digital synchronous sampling algorithm are successfully applied to commercial energy metering SoC design.