Integrating Quantum LSTM and Quantum Detector for Forecasting and Anomaly Detection for Load Data in Low-Voltage Power Distribution Networks

In the ongoing effort to tackle losses within the low-voltage power distribution network, a critical necessity emerges: the development of robust methodologies for forecasting and assessing data acquired from low-voltage measurement points. These methodologies serve as indispensable inputs for addressing the overarching challenge of minimizing grid losses. This paper presents a novel approach that leverages Quantum Long Short-Term Memory (LSTM) networks for predicting power consumption data, alongside a Quantum Detector for identifying and categorizing anomalies within the power data collected from these measurement points. The chosen algorithm is applied to a dataset comprising power (P) readings obtained from a representative measurement point within the Ho Chi Minh City power grid. Through extensive evaluation of the forecasting and data classification performance, the efficacy and viability of the proposed methodology are demonstrated. These findings underscore its potential for seamless integration into grid analysis frameworks, thereby facilitating subsequent endeavors aimed at mitigating power losses effectively.