Research On Transient Power Quality Detection And Monitor Data Compression Method

With the continuous progress of science and technology, such as semiconductor rectifierload in Modern circuit systems are increased. The nonlinear, impact and imbalance electricalcharacteristics of these electricity load, which cause increasingly serious problem of powerquality. Power quality monitoring is a direct way to acquire all kinds of power qualityinformation, and provide scientific basis for accurate, teal-time monitoring of power quality,so it has very important significance. Aiming at the problems of power quality monitoring,this paper mainly studies the transient power quality disturbance detection and three-phasequality monitoring data compression method.The paper first builds models of the voltage swell, voltage interruption, voltage sag,transient impulse and transient oscillation of five kinds of common transient signals in Matlabsimulation platform, and the output of various kinds of power quality signal for simulationanalysis.Aiming at the problem that it is difficult to fast, accurate detection and identification ofthe transient power quality, a new detection and classification method is presented based onwavelet transform and BP neural network. Based on wavelet modulus maxima detectionsignal singularity theory, performs multi-scale wavelet decomposition of disturbance signals,and then use the first stage high frequency coefficient by reconstruction to accurately positionthe disturbance of the start and end time, realized transient power quality disturbancesaccurately detection and time orientation. Using the disturbance signals energy distributionwhich formatted by the multi-scale wavelet transform coefficient as the feature vectors. Thenthe feature vector is input to the BP neural network. BP neural network are used to classifythese eigenvectors of different power quality disturbances. The simulation results demonstratethe feasibility of the method.At present, the power quality monitoring data has increased considerably; its storage andtransmission become a huge burden. To handle this problem, the paper introduces abc-dq0transform, and presents a compression method of the power quality monitoring data based onimage compression technology. First use abc-dq0transforms to eliminate the redundancy of the three-phase data, and then the one-dimensional power quality data are transformed totwo-dimensional matrix to eliminate the redundancy between cycles. To process the2Dimages using image smoothing technology. Meanwhile, process two-dimensional data base ontwo-dimension wavelet transform and image set partitioning in hierarchical tree codingalgorithm. The simulation results demonstrate the method has good compression performance,and can flexibly regulate the compression ratio.