Research On Detection Of Arc Fault In Series Connection Of Landscape Power Supply

With the development of urban construction,landscape lighting has been used on a large scale,and its safety has also become the focus of attention.Due to the long power distribution circuit of the landscape power supply,the circuit impedance is large and the fault current is small when a fault occurs,and the fault current is not enough to cause the protection of the circuit breaker.The specific situation analysis and discussion can determine the standard of its failure.In this paper,the feasibility of the proposed wavelet denoising and multi-feature fusion algorithm is verified by developing a high-power landscape power supply and a fault detection device.This thesis firstly introduces the development status of landscape power supply and the generation and harm of series fault arc,and then designs the overall hardware structure of the system,which includes two parts: landscape power supply and fault arc detection device.The main circuit of landscape power supply adopts LLC half bridge.Resonant circuit,the ZCS and ZVS it can achieve reduce the switching loss of the switch tube and the power consumption during reverse recovery of the diode,so that the efficiency of the power supply is improved;the fault detection part collects the current in the line through the current sensor,and after conditioning Enter the main chip for algorithm analysis to determine whether the fault arc occurs.Secondly,the simulation mathematical model of the series arc fault arc is established.The fault current waveform obtained by the simulation is helpful for us to analyze its fault characteristics more intuitively.At the same time,the simulation model of the driving power supply of the landscape lamp is built,which is combined with the arc fault simulation model to discuss The characteristics of the series fault arc in a specific scenario provide data support for the selection of the algorithm.Since there is a lot of high-frequency noise in the line in practice,it is necessary to filter the interference information in the signal through wavelet denoising to make the fault feature more obvious.After the denoising is completed,the fault feature quantity in the signal is extracted.The inter-harmonic mean value,discrete coefficient and wavelet singular entropy are fused to solve the limitation of single feature quantity for fault identification,and the comprehensive judgment standard for judging fault occurrence is obtained through the fusion of multiple feature quantities.Finally,the algorithm is verified by the developed physical platform.In this thesis,the series arc fault arc in the landscape power line is simulated and analyzed,and the algorithm of wavelet denoising and multi-feature fusion is proposed to detect the fault.This research result provides a more reliable technical guarantee for the safe use of landscape power.