An Intelligent Vector Control Technique for Autonomous Solar Photovoltaic Water Pumping System with Reduced Sensors

Typical vector control approaches extensively deployed for controlling standalone converters have limitations in dynamic conditions. Real-time handling of non-linear constraints and parameter tuning are major drawbacks for such sophisticated control approaches. Moreover, traditional solar PV water pumping systems utilize additional current sensors, which reduces reliability and increases the cost of maximum power point tracking systems. To overcome these limitations, this study presents a deep neural network-based vector control approach for an induction motor-driven solar PV water pumping system with reduced sensors. The framework explicitly considers source uncertainties, pump flow rate, current, and speed to improve pump operation and reduce energy consumption and cost. The proposed approach reduces the requirement for an intermediate power conversion stage between the dc-bus and PV array. Results obtained from the proposed approach adoption are found to be promising and motivate future research in this field. The approach will facilitate water distribution systems operating with more informed decisions, and work towards more sustainable system operation.