Integrated Tool for Cleaning Bulk Solar Power Data

Cleaning of photovoltaic (PV) power generation data is an important step during data pre-processing stage in a data analytics program. Scope of solar PV generation data analysis include solar power generation forecasting, energy management, optimal design of solar plants etc. Machine Learning applications in power sector are emerging rapidly in recent years. The intermittent nature of the output power of PV systems makes it difficult for a data engineer to pre-process the erroneous solar PV generation data. The proposed data cleansing algorithm aims to help the data engineer without domain expertise to process exclusive solar PV generation data. Generalized data cleansing tools like R and Python, holds many inbuilt tools to clean bulk data set. A machine learning based approach have been proposed to replace the missing values of the solar power and energy data set, with more suitable values, thereby minimizing the intervention of a solar data expert. Finally, the performance analysis of the developed algorithm is carried out using bulk solar power data collected from a roof top solar plant.