| Summary: | 碩士 === 國立臺灣科技大學 === 電機工程系 === 105 === The characteristic curve of a solar cell depends on its voltage and the current relation, which is changing with the illumination and the temperature. Owing to different conditions, each solar cell has different voltage-current (V-I) characteristic curve exhibiting different maximum power point. Therefore, maximum power point tracking (MPPT) system must be developed for tracking the maximum power under different shading conditions.
Due to partial shading effects and module deterioration, the efficiency of the traditional centralized solar system is extremely low. Therefore, people have proposed to use distributed photovoltaic system (DPS). Since in DPS, each panel will only exhibit a single-peak P-V characteristic, deterministic algorithms are sufficient for MPPT tracking. In this thesis, we have compared and studied five commonly used distributed MPPT (DMPPT), and they are Perturb and Observe (P&O), Incremental conductance (INC), Polynomial Interpolation (including Newton and Lagrange), and Fibonacci search (FS).
In order to compare their tracking performance, various partial shading conditions, extracted from EN50530, were emulated by a PV simulator, which was fed to an interleaved boost converter. Based on the experimental result, in most of the cases, Newton’s method outperformed the other methods in terms of tracking time and MPPT efficiency.
|
|---|
