Neural Network Based Source Selection Scheme for Wind-solar Based Auxiliary Supply in Railway Traction Systems

Abstract

The concept of catering electric supply to the auxiliary loads of railway wagons using Wind-solar based renewable energy systems is under consideration at number of metro rail projects. In such systems the selection of source of supply from available sources is a matter of concern. The existing schemes of source selection are mostly hardware oriented that invite problems such as delay in selection, components becoming sluggish and insensitive and loose connections. This paper suggests a novice neural network based approach for source selection as a replacement against the existing schemes. The proposed method of software abstraction is simple, low cost and accurate.