A linearised PV maximum power tracking controller for AC network integration analysis

Abstract

The maximum power available from a photovoltaic (PV) generator in many grid connected systems is extracted using a dc-dc step up converter that implements a maximum power point tracking (MPPT) algorithm. Traditional MPPT algorithms are iterative, continuously searching for the maximum power point (MPP) under varying weather conditions. Due to the nonlinear time-varying nature of commonly used MPPT methods such as perturb and observe (P&O) it is not easily integrated into state-space models for AC network analysis. In this paper a linearised PV MPPT controller for AC network integration analysis is presented. The non-linear step up converter and PV array are transformed into a time invariant system by using the state space sliding averaging technique, over a single switching cycle. This allows for linear controller design and provides a computationally inexpensive model suitable for AC network integration analysis. The linearised controller robustness was assessed over the expected operating irradiance range, at a constant temperature, using the detailed switching model in MATLAB/Simulink which produced increased performance and stability when compared to P&O MPPT. The computational time using a standard operating system for a 2s simulation of the detailed switching model was 263s while the reduced small signal model was 0.4s.