A rule based fuzzy model for the prediction of daily solar radiation

The main goal of this investigation is to use the fuzzy systems of Takagi Sugeno for the modelling of the solar radiation data. Generally, the process of identifying a fuzzy inference system (or fuzzy model) requires two types of tuning designated as structural and parametric tunings. The first one concerns the structure of the rules and deals with problems such as the partition of the universe of discourse, the number of fuzzy if-then rules and the number of membership functions for each input. The second one is the identification of the parameters of the system. In this work we used the fuzzy clustering techniques to determine the adequate structure, and the weighted least square (WLS) algorithm to estimate the linear parameters. To verify the effectiveness of the proposed fuzzy modelling method, the identified TS fuzzy model is applied to predict the global solar radiation data. The numerical results are then compared with the results of a model using the SOS techniques. It is shown that the fuzzy modelling approach is not only more accurate than the SOS techniques but also provides some qualitative information.