Design Of Capacitor Fault On-line Monitoring And Residual Life Prediction System

Capacitor as the important reactive power in electric power system, these actual working conditions will seriously affect the reactive power balance in power grid.But the working environment of capacitors in the power grid is complex, this reason causing capacitor fault frequently, and the actual working life compare with the theory life varies greatly. Capacitor fault impact the safety of power grid and reliability of power supply. However, the fault detection methods for capacitor is still using the power-off detection, this method can not reflect the actual working state of the capacitor, and also cause a large number of waste such as various resources, and the worst result is line outages. In view of this situation, this thesis select on-line monitoring of capacitor fault and remaining life as the research content.Based on the analysis the electrical equipment and methods of the prediction of the life expectancy, added the means of fault diagnosis of capacitance equipment.This thesis puts forward the method of on-line monitoring of capacitor fault in combination with the national standard. And then design a capacitor life prediction scheme based on BP neural network algorithm. Then, the thesis analyzes the capacitance value of capacitor and the data related to the life value of capacitor value by using the MATLAB neural network toolbox. Based on the capacitance value. And then the TMS320F28335 chip is used as the controller of the system to design the hardware system, and the software system is designed in the CCS environment.Finally, under the laboratory environment, the designed capacitor fault monitoring and residual life prediction system debugging test, mainly including signal acquisition, data analysis and alarm module testing, testing the feasibility of the design program.