As the core component of photovoltaic grid-connected system,the performance improvement of inverter is an important means to improve the performance of photovoltaic grid-connected system. As the key components of the inverter system,capacitors and inductors play the role of energy storage and filtering in the system,and their electrical characteristics will directly affect the dynamic and static characteristics of the photovoltaic inverter system.The existing research showed that capacitors and inductors should be fractional in nature,and the modeling,analysis and control of inverter systems were carried out on the theoretical basis of integer calculus,which will inevitably produce some errors.Therefore,the fractional controller of three-phase photovoltaic inverter system based on the improved Oustaloup algorithm is introduced into the grid-connected photovoltaic inverter system in this paper for fractional modeling, analysis and control,which can provide a theoretical model for the simulation and analysis of fractional power electronic systems.The practical operation in Jining PV power generation base (installed capacity of 560MW) in Shandong Province,China,showed that the fractional-order PI ^λ controller had a good control effect on the grid-connected PV system,and the power quality of grid-connected PV was also improved.

Virtual reality (VR) technology provides specific three-dimensional (3D) scenes for users, which could be benefit for the applications in the field of medical diagnosis, psychological analysis, cognitive researches and entertainments. The current VR device provides the virtual 3D images, but cannot synchronously detect the user's eye movements which could be significant for deriving the users' gaze points and interest regions in varied applications. In this paper, we proposed a smart eye tracking system for VR device. Firstly, a smart eye movement detection kit is mounted inside the VR device to capture the human eye movement images with the rate of 30 Hz. Based on the proposed eye detection kit, a gaze detection algorithm is then established for VR devices. Next, a gaze estimation model is proposed for human gaze estimation. The proposed smart eye tracking system can detect the user's eye movements in real time under VR stimulation. Moreover, it also provide a new human-VR interaction mode.