maximum power point tracking in photovoltaic systems using model reference adaptive control

摘要

This thesis proposes adaptive control architecture for maximum power point tracking (MPPT) in photovoltaic systems. Photovoltaic systems provide promising ways to generate clean electric power. MPPT technologies have been used in photovoltaic systems to deliver the maximum power output to the load under changes of solar insolation and solar panel's temperature. To improve the performance of MPPT, this thesis proposes a two-layer adaptive control architecture that can effectively handle the uncertainties and perturbations in the photovoltaic systems and the environment. The first layer of control is ripple correlation control (RCC), and the second layer is model reference adaptive control (MRAC). By decoupling these two control algorithms, the control system achieves the maximum power point tracking with shorter time constants and overall system stability. To track the maximum power point as the solar insolation changes, the RCC algorithm computes the corresponding duty cycle, which serves as the input to the MRAC layer. Then the MRAC algorithm compensates the under-damped characteristics of the power conversion system: the original transfer function of the power conversion system has time-varying parameters, and its step response contains oscillatory transients that vanish slowly. Using the Lyapunov approach, an adaption law of the controller is derived for the MRAC system to eliminate the under-damped modes in power conversion.