A New Design Method for Solar Energy Harvesting System Based on Neural Network

This paper presents a new design method based on neural network and co-simulation of MATLAB/Simulink and Virtuoso for solar energy harvesting system to power IoT applications. With the help of this new design method, a solar energy harvesting system with MPPT is designed and fabricated in 0.18µm CMOS technology. The proposed MPPT scheme was able to track the maximum power point successfully with average tracking errors of 0.30% (pre-simulation) and 0.42% (test). And the efficiency is about 72.9%∼76.8% when the light intensity is changing from 5000lux to 11000lux. In summary, the proposed new design method has the advantages of high accuracy, controllability and operability.