Research Of Localization Of Electric Spark Discharge Source

With the development of society, people’s dependence on electricity is growing, the number of electric sparks occur also increases. Spark occurs with physical properties such as sound, light, heat and electromagnetic wave. They are not only interfering with the normal work of the device, even bringing about serious damage to the equipment or causing a fire. When the fire occurs, it will bring not only economic loss but also result in casualties, poses a great threat to the daily production and living. Given the occurrence of electric spark cannot avoid completely, so the monitoring of electric spark location has great practical significance. Locating the position of the spark when a spark occurs can reduce the loss before the spark brings great harm.This paper analyzes the formation and the characteristics of electromagnetic wave of the electric spark, introduces the equivalent form and the measuring method of electric spark. The author analyses and simulates the discharge voltage waveform characteristics, establishes electrical spark discharge electromagnetic model based on electric dipole model and improved electrical dipole model.The author summarizes the electric spark location method, introduces the method of two-dimensional spatial location, and deduces the three-dimensional space positioning method based on wireless sensor network location technology(WSN), selects the two positioning methods to achieve electrical spark sources positioning research.In the two-dimensional space, the author selects particle swarm optimization(PSO) algorithm as the optimization method, establishes the magnetic field localization model and objective function of electric spark source, uses PSO algorithm to simulate and locate the source of the spark. Under the assumption that environment, the better positioning accuracy is obtained by MATLAB simulation; then in the laboratory environment, doing the actual spark source localization experiments, selecting six detection sensors to collect data for the PSO algorithm. In the end, the positioning error is less than 0.06 m, verifying the validity of PSO for the spark source localization.In the three-dimensional space, the paper selects the improved weighted centroid algorithm for spark source localization research. He makes theoretical derivation and experiment of variation trend of discharge voltage, and then gets the ranging model of electric spark discharge voltage. The author analyzes the detection node layout and improves the weighted centroid algorithm. In the laboratory environment doing the spark source localization experiments, the average distance error of several positioning experiments is 0.265 m, got a positioning results, verifying the validity of the improved weighted centroid localization algorithm.