Modelling of a directly coupled PV water pumping system using soft computing techniques

Abstract

Photo-Voltaic Water Pumping System (PVWPS) is the most suitable system for low head irrigation in the remote areas. PVWPS may be characterized by its multivariable-nonlinear equations; however, efficient water demand management necessitates fast and accurate water flow rate estimation at actual operating states. This can hardly be achieved with conventional mathematical -based methodologies. In this paper two of the Soft Computing (SC) techniques are due selected; Artificial Neural Network (ANN) and Adaptive Neuro-Fuzzy Inference System (ANFIS) for PVWPS novel modeling and performance pre-evaluation .The (ANN) model and (ANFIS) models are trained off-line to identify the water flow rate based on air temperature, solar irradiation, and static head as input parameters. An iteration technique which is mathematical-based is also presented for comparison and evaluation purposes. The paper indicates accuracy, robustness and effectiveness of the proposed models in water flow rate control, economic feasibility and fault detection.