A power control system in DC-DC boost converter integrated with photovoltaic arrays using optimized back propagation Artificial Neural Network

This paper presents a power control system for DC-DC boost converter integrated with Photovoltaic (PV) arrays using optimized back propagation Artificial Neural Network (ANN). A specific output voltage of a boost converter is regulated at a constant value under input voltage variations caused by sudden changes in ambient temperature and irradiation for a purpose of charging batteries in stand-alone solar power systems. The back propagation ANN controller is realized for regulating an output voltage. The optimization of ANN topologies is also studied in order to achieve an optimal structure that suits for simple hardware implementations, and exhibits fast setting time with low overshooting. Simulations have been performed in a computer-aided design tool MATLAB Simulink using a commercially available PV array model: Kyocera KD135GX-LP. Test conditions include step changes in ambient temperatures varied from 10°C to 45°C and irradiations varied from 200W/m2 to 1000W/m2. Under a particular operating condition of 12V to 24V voltage conversion and a 50Ω resistive load, the output power variations under such step changes are less than 0.2W. The system possesses fast settling time of 6.4ms with low voltage ripples of approximately 0.625%. With a complete consideration of changes in environment conditions, this paper offers a potential alternative to robust output power control in modern PV power generation systems.