Background:
Despite so many developments, most of the farmers in the rural areas are
still dependent on rainwater, rivers or water wells, for irrigation, drinking water etc. The main reason
behind such dependency is non-connectivity with the National grid and thus unavailability of
electricity. To extract the maximum power from solar photovoltaic (SPV) based system, implementation
of Maximum Power Point Tracking (MPPT) is mandatory. PV power is intermittent in nature.
Variation in the irradiation level due to partial shading or mismatching phenomena leads to the development
of modular DC-DC converters.
Methods:
A stand-alone Multi-Input Dual-Output (MIDO) DC-DC converter based SPV system, is
installed at a farm; surrounded with plants for water pumping with stable flow (not pulsating) along
with battery energy storage (BES) for lighting. The proposed work has two main objectives; first to
maximize the available PV power under shadowing and mismatching condition in case of series/
parallel connected PV modules and second is to improve the utilization of available PV energy
with dual loads connected to it. Implementation of proposed MIDO converter along with BES addresses
these objectives. First, MIDO controller ensures the MPPT operation of the SPV system to
extract maximum power even under partial shading condition and second, controls the power supplied
to the motor-pump system and BES. The proposed system is simulated in MATLAB/ SIMULINK
environment. Real-time experimental readings under natural sun irradiance through hardware
set-up are also taken under dynamic field conditions to validate the performance.
Results and Conclusion:
The inherent advantage of individual MPPT of each PV source in MIDO
configuration, under varying shadow patterns due to surrounding plants and trees is added to common
DC bus and therefore provides a better impact on PV power extraction as compared to conventional
PV based water pumping system. Multi-outputs at different supply voltages is another flag of
MIDO system. Both these aspects are implemented and working successfully at 92.75% efficiency.
