|
Photovoltaics generate electricity directly from sunlight via solar panels, but Taipower can only receive AC current, so it must convert the DC current into AC current through an inverter device. The Inverter will aging to fail. When the inverter cannot generate electricity, it is found that the total power generation of inverter has not changed. The system will judge it as a failure, and will result in a loss of revenue. Therefore, how to effectively diagnose the fault of the inverter is worth exploring. This research develops a diagnosis system for solar inverter, in order to estimate the failure time of the inverter. In terms of the fault diagnosis mechanism, the real-time power data captured by solar inverter is used to calculate the AC and DC values of the inverter, analyze the coefficient of variation for fault prediction. When an abnormality is found, the prediction system immediately informs the manager, allow the manager to make follow-up dispatch and equipment scheduling. The research results show that through the simulated AC and DC values of the converter, the analyzed slope characteristic of the coefficient of variation has risen for 8 consecutive days. It can be judged that the inverter is about to fail and stop power generation, and the system can be predicted in advance. The warning message informs the manager to reduce the waiting time for maintenance by more than 2 weeks. Therefore, the solar converter diagnosis system designed by this research institute can achieve the purpose of diagnosing the fault of the inverter.
|