DESIGN & IMPLEMENTATION OF MPPT SOLAR PHOTOVOLTAIC - ELECTRIC VEHICLES IN FAST VARYING PARTIAL SHADING CONDITIONS USING SERVAL OPTIMIZATION ALGORITHM

Abstract

Solar panels are used to convert solar energy into electrical energy. In this study it was applied to electric vehicles which have a very large potential for being constrained by shadows. In fast varying partial shading conditions, the position of the maximum power point is divided into two, namely GMPP and LMPP. This condition makes the MPPT process stuck in LMPP. Therefore, this research proposes the application of Serval Optimization Algorithm (SOA) in MPPT. This method refers to the natural behavior of the serval in nature. The fundamental inspiration of SOA is the serval hunting strategy in two stages of exploration and exploitation. The SOA is implemented in MPPT to change (duty cycle) so that it gets the best value and produces maximum solar panel output power. This SOA method was chosen to complete the partial shading conditions so that MPPT can optimally reach GMPP without going through LMPP. The solar panels used in this system are 2  units with specifications of 25 Wp with a 24V battery load and a 120W BLDC Motor compact in Electric Vehicle - Two Wheeler Scooter. MPPT SOA was tested in a simulation using PSIM and actual Software in 6 variations of normal and partial shading conditions. In the Simulation Test of Partial Shading Conditions, an average accuracy of 99.958% and an average tracking time of 0.492 seconds were obtained. SOA has a higher accuracy than PSO and GWO, which is 99.95%. And it has a faster tracking time of 0.55 seconds. In the SOC Integration Test, the SOA Method obtained an error value of 6.48% better than the GWO Method. On the Road Test with 6 condition, it can slow down the value of the decrease in battery capacity by 16.24%. The application of Single Source on the PV-MPPT-Converter can be implemented with an efficiency value of 76.86%. In previous research where SOA is a new method in Optimizing Problem Solving which has quite good accuracy performance, and in this research it can be implemented in Solar PV Optimization to track MPPT electric vehicles with varied and fluctuating partial shading conditions.