Development of performance index for evaluation of small scale hydro power plants by neural network and multi criteria decision making

In recent years the small scale hydro-power projects have emerged as a viable and less-expensive alternative to conventional energy sources. The suitability of such projects in a given location must be analyzed based on site-specific factors, and this is often performed with the assistance of MCDM methods, of which, FLDM models are the widely applied. However, both methods have their drawbacks. FLDM is known for its 'haziness' when converting its fuzzy rating into a crisp rating, while random selection of weight vector for attributes becomes a major problem and ANN to predict the index through the weight function. If fuzzy logic is used to determine the weight vector to be assigned to the criteria considered for a certain decision-making problem, the output of the result will be more logical and the haziness of the conversion to a crisp rating will not influence the decision. Thus, we investigated a hybrid FLDM method to identify the most suitable location for a small scale hydro-power project. The index also provided a heuristic and cognitive optimal value to way from a suitability of small scale hydro power plant installation. Both models were able to fit the data well, with R-2 values of 0.994074561 and 0.9964 for the linear regression model and the ANN model respectively. It was also found that the test dataset had a mean squared error of 0.0337 for the ANN model, while it was 0.03898 for the regression model.