Experimental Investigation to obtain the PV Module True Performance and Emphasis on Application of Soft Computing
Abstract
With the increasing interest in generation of electricity from Solar Photovoltaic systems, it has become important to evaluate true performance. One of the advantages of renewable energy system that these are distributed in nature and can generate near the load thus almost negligible transmission and distribution losses but another side they suffer from other types of losses so the performance is yet questionable. It is well proven that Solar PV system performance deteriorates , when deployed in natural conditions although manufacturer indicate the efficiency behind the module in STC ( 1000 W/m2 , 250C and A.M. 1.5) but it is difficult to operate at standard efficiency.
In this paper attempts are made to observe the natural wind effect on temperature of PV module and to evaluate the module performance by comparing different methods with experimental and simulated values. In order to predict the performance near true value, a neural network based back propagation method is used. This paper will be helpful for project planners to estimate accurate size and performance; with the help of neural network, they can estimate true potential before actual installation. It is noted that considering natural wind cooling effect net energy gain is about 25 Wh and various types of errors like mean square error(MSE) , root mean square error (RMSE), Mean bias error (MBE) and Standard deviation error (SDE) are less for cooled module than not cooled module.
