198 / 2021-10-31 19:21:07

Study on artificial immune mechanism for distribution network fault diagnosis

Power system and automation > Power system fault simulation and risk assessment

Single-phase ground faults are the most common faults in power distribution systems. It mostly occurs in humid and rainy weather, caused by many factors such as tree barriers, single-phase breakdown of insulators on power distribution lines, single-phase disconnection, and small animal hazards. Single-phase grounding not only affects the user's normal power supply, but also may produce overvoltage, burn out the equipment, and even cause a phase-to-phase short circuit to expand the accident. The 10-35kV line of China's urban power grid is mainly a neutral point non-effectively grounded system. After a single-phase grounding fault occurs, the fault characteristic signal is weak and difficult to extract, making accurate fault diagnosis difficult. In recent years, the power system has attached great importance to the reliable operation of the distribution network. More than 100 power distribution automation demonstration projects have been built with fault diagnosis and isolation functions. Since 2015, efforts have been intensified to promote the application of distribution automation and improve the ability to handle faults. The development of feeder automation and intelligent distributed distribution automation standards and demonstration applications have been carried out, and the ability to deal with weak characteristics such as single-phase grounding has been strengthened. However, due to the limitation of the traditional fault diagnosis principle that relies on the zero sequence voltage and current criterion, the current single-phase ground fault diagnosis in the distribution network has serious misjudgment and missed judgments. Taking single-phase ground fault diagnosis as an example, although the accuracy rate can reach more than 90% or even close to 100% under the experimental conditions (from the products of multiple manufacturers and the scientific and technological project indicators of the State Grid Corporation), the accuracy rate in practical applications Only 30-40%. With the rapid development of the distribution network in recent years, the functions such as wave recording and event playback of the distribution automation system and its equipment can collect online grid operation information before and after the accident, forming a sample database of accident cases. How to use these data to improve the accuracy of fault diagnosis will be the focus of future research. Artificial immunity has not been effectively carried out in the research and application of distribution network. Immune technology can use the recognition, memory, learning, and response mechanisms of immune carriers to modify risk warning and fault diagnosis models through the analysis of ever-increasing event samples, which is of great significance for improving the applicability and accuracy of practical applications.