Research On Low-voltage Arc Fault And Diagnostic Method

With the increasing number of household electricity, the threats of the electric fire accidents become more and more serious. When an arc fault occurs, the arc current can produce local high temperature and may cause fires if combustible materials surrounded. As conventional circuit breaker or other protection devices are unable to provide comprehensive protection for arc fault, study on the characteristics of low-voltage arc fault and proposing a recognition method is of great significance.In China, study on low-voltage arc fault is still at the starting stage. In order to propose the diagnosis method for arc fault, which is suitable for our country's residential electricity, the research work accomplished is shown as follows:First, experiments are conducted under AC 220V voltage (50Hz) according to UL1699 standard. Several typical loads are chose to do a number of experiments under different conditions and data are recorded.Then clustering is visualized using SOM neural network and K-means cluster. According to clustering results, some typical characters of arc fault are found: "Current zero off" phenomenon, positive and negative half-wave asymmetry, aperiodic waveform and high levels of high-frequency harmonics. At the same time the arc fault diagnosis program is put forward.Finally, the Arc Fault Circuit Interrupter (AFCI) is designed, and a prototype is produced. The arc fault diagnosis program is verified by experiment based on the test of the prototype. The test result shows that the prototype can act immediately after an arc fault occurs, which is in line with the desired objective of this paper.In this paper, the research work is based on experimental data. The arc fault diagnosis method proposed is proved to be reliable and practical by testing, and it may provide directive significance and application value to the protection for the arc fault.