Investigation of Energy Management Strategies for Photovoltaic Systems - A Predictive Control Algorithm

The high cost of energy from stand-alone photovoltaic (PV) systems dictates that the available energy is used in the most efficient manner possible. To do so requires careful attention to the operating characteristics of all the components of the PV system and the operating decisions of system control. By applying an energy management approach to system control, over-all system operating performance may be improved, yielding either more services provided for a given system rating or a smaller system size for the same services. This paper reports progress toward formulating energy management strategies for stand-alone PV systems, developing an analytical tool that can be used to investigate these strategies, applying this tool to determine the proper control algorithms and control variables for a range of applications, and quantifying the relative performance and economics when compared to systems that do not apply energy management. Although the formulation was originally driven by stand-alone PV system needs, the results are also applicable to hybrid or grid connected systems. An example of a load system for which an energy management stategy is usable is a photovoltaic village power system with battery storage, water pumping with a water tank, lighting, and a differable domestic load. In this case there are three "storage" elements; the battery, the water tank and the differable load. The energy management strategy determines how the energy is routed to the water pump and domestic load and in and out of the battery.