Power Quality Event Classification with CNN-FFNN Embedding Fusion Method

The widespread application of grid-connected renewable sources, electric vehicle charging systems, causes power quality disturbances (PQD) on AC sources. The PQD affects the home appliances and also more in billing amount. The detection steps consist of data acquisition, pre-processing, detection and decision making. The execution of all the processes is a complex task and a suitable embedded processor is required. In this work, a Raspberry Pi board with ADC and signal conditioning circuit is used to acquire the data. All the pre-processing has been carried out using Python in real-time. In this study, an innovative approach for power quality event classification utilizing Convolutional Neural Network (CNN) and Feedforward Neural Network (FFNN) embedding fusion is proposed. The fusion of FFNN with CNN enhances the model's capability to capture spatial and temporal dependencies, leading to robust classification outcomes. After processing the events also will be monitored in the Google Firebase cloud.