Research And Design On Arc Fault Detection Devices

With the rapid development of our national economy, our consumption of production and residential electricity is increasing. At the same time the electric fire accidents are more frequent, including electrical fires caused by arcing faults which are occupying creasingly high proportion. The protection of the arc fault has caused wide public concern. Because the arc current is lower than that of overcurrent protective device setting when the arc fault occurs, and without earth leakage, thus traditional protecting appliances cannot provide effective protection of fault arc. Therefore, the research on the characteristics of arc fault and on the basis of development has a very important significance for the arc fault detection devices. The arc fault detection devices has been researched and designed in this article.Firstly, this thesis analyzes the current situation and development trend of domestic and foreign arc fault detection devices, which has been defined the topics of significance clearly. According to the technical specifications and requirements analysis, the overall design of intelligent evacuation indicating system has been given in this article, then the overall plan of hardware and software also has been designed. Finally, the paper has made the hardware chip selection and the software development environment elaborate.Secondary, algorithm of arc fault is studied in this paper, discrimination based on feature extraction and BP neural network algorithm which can be applied to the actual effectively. Then, the algorithm is tested in Matlab.Newly, the hardware of arc fault detection devices has been designed in this paper, including core control circuit, power supply circuit, current data acquisition circuit, communication interface circuits, sound and light alarm circuit, key test circuit, external expansion interface circuit, etc., and details the principle and concrete realization of each part of the circuit. The software of arc fault detection devices has been designed in this paper, including main program, initialization program, A/D conversion process, feature extraction procedures, BP neural network training program, BP neural network assessment procedures, communications and sound and light alarm procedures, and the work process and program flow are given.Finally, the arc fault detection device has been tested with series arc fault, parallel arc fault, mutual interference, failure alarm. Test results show that the samples of arc fault detection devieces in this paper, achieve the desired design goals.