According to the requirements to pd pattern auto - recognition , this paper studies systematically the basic theories and realizable methods for auto - recognition of pd gray intensity image : ( 1 ) in the requirement of on - line pd monitoring for transformer , several discharge models are designed and the relevant experiment methods projected . with discharge model tests , a lot of discharge sample data is acquired . on the base of systematical research on recognition for pd gray intensity image , this paper puts forward two kinds of fractal features , the 2nd generalized dimensions of original pd images and fractal dimensions of high gray intensity pd images , and then the relevant extraction methods 針對局部放電模式自動識別的需要,作者系統(tǒng)地研究了局部放電灰度圖像自動識別中的基本理論和實現(xiàn)方法: ( 1 )根據(jù)變壓器局部放電在線監(jiān)測的要求,設(shè)計了放電模型和實驗方法,并通過模型實驗獲得了大量放電樣本數(shù)據(jù),為構(gòu)造局部放電灰度圖像和采用bpnn進行識別作好準(zhǔn)備; ( 2 )研究了局部放電灰度圖像的構(gòu)造方法以及降維構(gòu)造32 32灰度和矩陣的方法;在用人工神經(jīng)網(wǎng)絡(luò)對局部放電進行模式識別時,分析了bp網(wǎng)絡(luò)的優(yōu)缺點,對典型bp網(wǎng)絡(luò)的結(jié)構(gòu)和學(xué)習(xí)訓(xùn)練算法提出了改進,采用帶有偏差單元的遞歸神經(jīng)網(wǎng)絡(luò)作為模式分類器;采用32 32灰度和矩陣進行bpnn識別結(jié)果表明這種方法是有效的。