Detection methods for common problems in solar home systems

Abstract

Solar power is a valuable source of electricity for users in the developing world, yet many solar home systems are working at marginal capacity or not functioning at all. This study has two purposes: (1) to determine how common problems afflicting these systems affect the voltage output, and (2) to use patterns in voltage as a means of detecting these problems, via diagnostic tools. Team SHINE simulated common problems on experimental systems, collecting voltage data from the batteries and panels. Using these data, we created computer algorithms to detect the problem affecting the system. After testing several detection methods, it was found that the most successful performed at 86.5% accuracy. The algorithms can be used in future research to create a device for detecting these problems, allowing them to be addressed earlier.