A Study on Arc Detection Technique Based on Low Frequency and High Frequency Signal Analysis Method
摘要
When a series arc occurs, noise is generated by the arc. In the frequency range below 30 MHz, it is difficult to distinguish between the normal state and the arc state. Therefore, the noise generated by the arc is extracted in the region free from interference by background noise, such as ambient and biological noise, through the LC filter. The extracted noise is amplified by the RF amplifier and converted into the Received Signal Strength Indicator (RSSI). When a series arc occurs, a shoulder section, where the current is zero, appears. Due to this shoulder section, the RSSI value exhibits a 120 Hz component when a series arc occurs. Additionally, the occurrence of a series arc leads to an imbalance in the current waveform and fluctuations in the current’s magnitude. This paper proposes an arc detection method based on the 120 Hz component increase in the RSSI and the change in the current magnitude. This paper proposes an arc detection method based on the increase in the 120 Hz component of the RSSI and the change in the current’s magnitude. The 120 Hz component of the RSSI is analyzed using the Goertzel algorithm, while the current’s magnitude change is used to assist in arc detection. Using the analyzed current and the 120 Hz component of the RSSI, this paper presents a series arc detection technique for AC systems with resistance, drill, and SMPS loads.