Field Experience Guidelines for Maintaining Passive Power Filters to Improve for Power Factor and Power Quality Using Deep Learning Algorithm

Abstract

With more than 20 years of field experiences in Passive Power Filters (PPFs) designs, operations, and maintenances in many large scale industries in Southeast Asia, this paper reveals a simplify yet robust technique to examine and evaluate the health of PPFs that may defect during their long term functioning to maintain power factor and power quality in the electrical system. Based on the long term experience of installing a large number of PPFs, it is found that to maintenance capacitor banks in a low temperature condition will be more benefit. Under this operation condition, not only the banks can control the %THD level within the range specified by IEEE 519-1992 and 2014 standard but also make the bank last longer. Finally, the method of inspection and maintenance are proposed using neutral network algorithms for deep learning with Python. The results are complete and can be applied in industrial plants very well.