Cooperative Extremum Seeking for Power Availability Detection of Photovoltaic Cluster

The power availability detection for geographically dispersed multiple Photovoltaic (PV) generators plays an important role in the management and control of PV-based smart grid systems. This paper proposed a distributed estimation scheme to enable local awareness of global PV capacity awareness at each PV station, using a cooperative extremum seeking conjecture. In particular, the proposed algorithm treated each PV as a node in networked systems, and takes full advantage of the discrete information flow among networked node, as well as model free behavior of extremum seeking method. More specifically, a dynamic average tracking algorithm is introduced to guarantee each PV not only be able to track its own individual maximum power point (MPP), but also the total MPP (i.e. power availability) of the PV cluster simultaneously and locally without any centralized facility. Simulation results validate the effectiveness of the proposed method.