A Comparative Analysis of Support Vector Machine and Decision Tree Algorithm for Predicting Fault in Uninterruptible Power Supply Systems

Abstract

Power supply systems can have problems, and Ghana Gas Limited is not an exception. Ghana Gas Limited uses an intricate Uninterruptible Power Supply (UPS) system which is made up of several parts such as electromechanical components, PCB boards, and electrolytic capacitors. The majority of components have technical lifespans that are governed by usage, operational environment, and working conditions, such as electrical stress, working hours, and working cycles. Most of the time, these errors affect the integrity and power supply after manufacture. The issue is that it takes longer for the professionals who operate on this machine to recognize these flaws, which makes it difficult for them to predict errors quickly or anticipate the likelihood of faults happening in the system components at an early stage for effective corrective action to be performed. Support vector machines (SVM) and decision trees were used in this study to anticipate faults for technical data scheduling of uninterruptible power supply systems for Ghana Gas Limited in an efficient manner. Based on a comparative analysis using these two techniques, faults in Ghana Gas Limited's power supply system were predicted using a four-hour daily interval dataset on UPS recordings, including input voltage, battery voltage, battery current, and alarm, spanning from August 2017 to October 2023. The findings depicted that the support vector machine was more efficient in detecting the fault locations in the power supply system with an accuracy of 96.80%, recall of 99.80%, precision of 100 %, F1-score of 93.15%. The results from the error metrics also validate the measures in assessing the predictive ability of the model with MAE of 0.42%, MSE of 1.18%, RMSE of 4.45%, R2 of 99.97%, RMSLE of 0.036%, and MAPE of 0.21%.