‘Self-learning Tracking’: a New Control Strategy for PV Concentrators
✍ Scribed by J. C. Arboiro; G. Sala
- Publisher
- John Wiley and Sons
- Year
- 1997
- Tongue
- English
- Weight
- 391 KB
- Volume
- 5
- Category
- Article
- ISSN
- 1062-7995
No coin nor oath required. For personal study only.
✦ Synopsis
Usually the tracking system is not given much importance when designing a photovoltaic (PV) concentrator, partly because the intensive work carried out on this subject has provided it with a false sense of maturity. However, only a few tracking systems have ever been successfully implemented in practical concentrators and many systems never worked in spite of the perfect theoretical design. In this paper we present a review of the experience in tracking systems from the Institute of Solar Energy. This experience runs in parallel with the evolution and development of such systems. It starts with the SANDIA experience in 1977 and moves on to discussing the problems and lessons learned with both open and closed loop systems in the RamoÂn Areces project (1978). After a brief description of self-aligning systems, we ®nish by discussing a new approach to tracking: the self-learning concept. In this case the tracking system monitors continuously the operation current, re®nes constantly its knowledge of the errors and misalignments aecting tracking accuracy and performs the corrections required. The algorithm for this scheme, which has already proved reliable for operation in the EUCLIDES concentrator prototype (1995 to date), is described in some depth.