Modeling and Simulation of MPPT Recurrent Neural Network (RNN) Under Partial Shading Condition Using CUK Converter

Solar energy is a renewable energy which has great potential to be developed in Indonesia. Solar energy can be converted to electrical energy using solar panel devices. Solar panels can get maximum power during sunny days, at optimum temperatures and the surface of the solar panel is not covered, but there is a possibility that the surface of the solar panel is covered by a barrier, for example, shadows from buildings, trees, or leaves that fall over the surface of the solar panel. This causes the power generated to be smaller so it cannot provide optimum energy. The partially shading solar panel surface conditions can also cause more than one maximum power (Maximum Power Point) on the characteristic curve P-V called GMPP (Global Maximum Power Point) and LMPP (Local Maximum Power Point). Conventional MPPT will most likely trapped in LMPP so that it cannot reach GMPP. Therefore, this paper proposes the Modeling and Simulation of MPPT RNN (Recurrent Neural Network) Under Partial Shading Conditions. The MPPT RNN is applied to the CUK converter for DC load. Based on the simulation result, known that MPPT RNN has higher accuracy when compared with MPPT P&O. This method resulted in 96.72% on partial shading conditions.