电能质量扰动在线辨识装置  

Device for on-line power quality disturbance recognition

作　　者：张明[1] 李开成[1] 

机构地区：[1]华中科技大学电气与电子工程学院,湖北武汉430074

出　　处：《电力自动化设备》2009年第9期124-128,共5页Electric Power Automation Equipment

摘　　要：设计了一种具有电能质量(PQ)扰动实时在线检测与分类功能的电能质量分析装置。该装置基于DSP和FPGA平台,实现了信号的采集、处理和显示。在算法上由于人工神经网络、专家系统、模糊逻辑、支持向量机等分类器过于复杂,故采用一种简单、高效的PQ扰动分类和量化方法,即基于规则基的决策树RBDT(Rule-Based Decision Tree)模式识别方法,同时提取5个典型的PQ扰动时频特征量作为决策树的输入量,实现了9种典型PQ扰动的辨识。通过算法仿真及硬件平台验证,结果表明可以满足对PQ扰动分类的精度和实时性的要求。A device for on-line detection and classification of PQ（Power Quality） disturbances is designed, which uses DSP,FPGA and simple peripheral circuit to realize the functions of signal acquisition,processing and display. In stead of the complicated classifiers,such as artificial neural network,expert system,fuzzy logic or support vector machine,a simple and effective method of PQ disturbance classification and quantification is proposed,which is a pattern recognition method of RBDT（Rule-Based Decision Tree） based on rule base. Five typical time-frequency characteristics of PQ disturbances are extracted as the inputs to recognize the pattern of nine typical PQ disturbances. Its performance is verified by algorithm simulation and HW test,showing that it meets the requirements of classification accuracy and real-time performance.

关 键 词：电能质量 扰动分类 在线辨识 DSP+FPGA 决策树 规则基 

分 类 号：TM712[电气工程—电力系统及自动化] TP277[自动化与计算机技术—检测技术与自动化装置]