Influence of A/D quantization in the fast and accurate grid signal estimation for a PV system using I class Rife-Vincent time windows

摘要

Solar energy is one of the most popular renewable energy sources. The control process of the photovoltaic system is usually based on the estimation of grid signal parameters. This paper presents an estimation method based on the FFT procedure, I class Rife-Vincent time windows and the 3-point spectrum interpolation. The accuracy in a real time system depends on several parameters: number of samples N, the value of window order H, number of A/D converter bits b, the value of signal periods in a measurement window CiR. Total estimation errors consist of two components but in practice (N = 64, ..., 2048; b = 8, ..., 24) the component caused by using the A/D converter is dominant (inversely proportional to the square root of the N value and 2b). Presented expressions allow to calculate maximum estimation errors for given values of b, N, H and CiR before starting the estimation process which is very important from a practical point of view (e.g. optimal selection of the A/D converter). The estimation errors for CiR = 0.5 (10 ms), H = 3, b = 16 and N = 512 are approximately 10 -5 Hz/Hz.