Currently, sizing and evaluating photovoltaic installations are based on average consumption profiles rather than scientific criteria or load models. This practice leads to installations fulfilling their tasks, but far from energy and economic optimization conditions that could make the photovoltaic option competitive, reliable, and attractive for more and more applications. Besides, a static knowledge of a photovoltaic system, as obtained during a purely energetic dimensioning, does not allow a reliable approach to the analysis of the system, nor the guarantee of the desired performances. Our work proposes to fill this gap by developing new tools for the simulation, sizing and control of photovoltaic systems. This study is based on three important points: - the development of a dynamic model for each photovoltaic application (pumping, electrification, refrigeration), - the improvement of the DC/DC and DC/AC conversion devices and this by the generation of new control laws associated with the MPPT and PWM modes- The integration of these techniques and these models in software for the management and control of the photovoltaic plant.