Among the different desalination systems, reverse osmosis is popular due to its specific characteristics. Two challenges in these systems are power supply and compensation for the imposed disturbances and uncertainties which affect the performance of the system. To study these two challenges all main subsystems of the photovoltaic-reverse osmosis (PV-RO) system are considered. A photovoltaic system is considered for power supply and robust controllers are proposed to deal with the uncertainties and disturbances. For these purposes, all parts of the PV-RO desalination process are considered. Maximum power is tracked in different sun radiations and temperatures using a Mamdani-type fuzzy controller optimized by the invasive weed algorithm (IWA). This new MPPT controller gets about 10 % more power, 60 % lesser fluctuations, and 20 % smaller rise time. The speed controller of the induction motor and the pressure controller are fuzzy-PID. An adaptive time-varying sliding mode control (ATVSMC) is proposed to control the RO system because this system is highly nonlinear with many uncertainties and disturbances. The controller has global robustness and regulates the system to the desired points at a specified time. In this controller, knowing the maximum limit of the system uncertainties is not essential, which is a practical advantage for the PV-RO system. The numerical simulations in different scenarios and a comparison to existing work show the superiority of the proposed controllers. This novel controller decreases the overshoot and settling time of the bypass stream velocity by about 3.5 % and 7.6 % in comparison to super-twisting sliding mode control (STSMC), respectively. The settling time of the retentate stream velocity has decreased by about 9 %. Moreover, there are no fluctuations in the control inputs and outputs, and the amplitude of the bypass stream control input has decreased by about 98 %.
Read full abstract