Title :
A biomimetic strategy for designing easily-installable CPV tracking system with high wind resistivity
Author :
Yamamoto, Manabu ; Futakuchi, Ryutaro ; Inoue, Ken ; Arase, Hidekazu ; Matsushita, Akira ; Itoh, A. ; Asano, Takashi ; Hayashi, Neisei ; Inoue, Daisuke ; Nakagawa, T. ; Fujii, Eiji ; Ueda, Daisuke
Author_Institution :
Adv. Technol. Res. Lab., Panasonic Corp., Kyoto, Japan
Abstract :
Due to its higher photovoltaic efficiency, CPV has the advantage of needing a smaller panel footprint than fixed PV systems. This characteristic makes it ideal for use as an independent power supply in remote locations where transportation and site procurement present a challenge. The overall system needs to be compact and light without sacrificing weather resistance. Skilled personnel are scarce in remote areas, so easy setup and maintenance are also important. We have developed two techniques based on the notion of bio-mimicry. One is machine learning to counteract errors in installation position without the need for human intervention, and the other is a three-dimensional panel structure that cuts the wind load by up to 20%. We describe our experimental results gained in field tests.
Keywords :
learning (artificial intelligence); photovoltaic power systems; power engineering computing; power supplies to apparatus; CPV tracking system; biomimetic strategy; concentrated photovoltaics system; independent power supply; installation position; machine learning; photovoltaic efficiency; remote locations; skilled personnel; three-dimensional panel structure; weather resistance; wind load; wind resistivity; Accuracy; Batteries; Kalman filters; Reliability; Sun; Wind; Concentrated photovoltaics (CPV); Kalman filtering; automatic learning; bio-mimicry; installation position error; staggered; tracking; wind force;
Conference_Titel :
Photovoltaic Specialists Conference (PVSC), 2013 IEEE 39th
Conference_Location :
Tampa, FL
DOI :
10.1109/PVSC.2013.6744485