Abstract:
This work on Photovoltaic pumping systems is composed of two broad sections. The first section presents an application of an Artificial neural network (ANN) for the•identification of optimal operating point of a PV supplied separately excited dc motor driving two different load torques. An ANN model is developed to control a directly coupled photovoltaic water pumping system. A gradient descent algorithm is used to train the ANN controller for the identification of maximum power point of the Solar Cell Array (SCA) and gross mechanical energy operation of the combined system. A non-adaptive off-line trained controller using ANN is developed for obtaining the MP (maximum power) and GME (gross mechanical energy) of the PV supplied motor-pump system. The model and algorithm is developed based on matching of the SCA to the motor load through a buck-boost power converter. Performance analysis is evaluated for the simulated results in MATLAB.
• The second section of this work consists of modeling a directly coupled photovoltaic pumping system using ANN to predict the pump flow rate for a given pumping system for different head values. A method for modelling the output of solar photovoltaic water pumping was adopted for training the neural network. This method relies on the data that can be quickly measured. Performance evaluation was done in terms of predicted output values of pump flow rate and percentage error in prediction. It also demonstrates the error of back propagation technique for the neural network.
