基于ECA-CNN的高壓隔離開關(guān)狀態(tài)識別方法
張咪,鮑哲,吳澤鋒,江鈺韜
(西安西電開關(guān)電氣有限公司,陜西 西安 710077)
摘 要 :為了解決高壓隔離開關(guān)狀態(tài)識別困難、實(shí)時(shí)監(jiān)測能力不足的問題,構(gòu)建了融合嵌入式感知和輕量化神經(jīng)網(wǎng)絡(luò)的智能監(jiān)測體系。硬件層采用模塊化設(shè)計(jì),智能感知單元與后臺控制系統(tǒng)形成閉環(huán)聯(lián)動,捕捉開關(guān)動作觸頭狀態(tài)關(guān)鍵特征。軟件層提出雙階段優(yōu)化策略:在數(shù)據(jù)預(yù)處理階段,構(gòu)建融合幾何 變換 ( 旋轉(zhuǎn) / 剪裁 )、直方圖均衡化、圖像濾波和噪聲注入等構(gòu)成的圖像特征增強(qiáng)算法,有效提升圖像 質(zhì)量并擴(kuò)充樣本多樣性;在特征提取階段,設(shè)計(jì)深度優(yōu)化的融合高效通道注意力 (ECA) 機(jī)制的卷積神經(jīng) 網(wǎng)絡(luò) (CNN) 模型,通過高效通道注意力機(jī)制模塊實(shí)現(xiàn)通道特征權(quán)重動態(tài)分配,強(qiáng)化關(guān)鍵特征提取能力。 算法分析驗(yàn)證表明,設(shè)計(jì)的圖像增強(qiáng)算法能夠有效提升圖像質(zhì)量,突出關(guān)鍵特征,使識別模型準(zhǔn)確率提升超30%;提出的ECA-CNN模型能夠在圖像增強(qiáng)算法的基礎(chǔ)上,進(jìn)一步提升對圖像關(guān)鍵特征的關(guān)注度, 其識別準(zhǔn)確率高達(dá) 97% 以上,較CNN模型提升12%。
關(guān)鍵詞 : 高壓隔離開關(guān) ;狀態(tài)識別 ;圖像增強(qiáng) ;卷積神經(jīng)網(wǎng)絡(luò) ;注意力機(jī)制
中圖分類號 :TM564.1 文獻(xiàn)標(biāo)識碼 :A 文章編號 :1007-3175(2025)12-0063-08
A State Recognition Method for High-Voltage Disconnect Switches Based on ECA-CNN
ZHANG Mi, BAO Zhe, WU Ze-feng, JIANG Yu-tao
(Xi’an XD Switchgear Electric Co., Ltd, Xi’an 710077, China)
Abstract: To overcome the challenges of state recognition and insufficient real-time monitoring capability in high-voltage disconnect switches, this study establishes an intelligent monitoring system integrating embedded sensing and lightweight neural networks. The hardware layer adopts modular design, where intelligent sensing units and background control systems form closed-loop linkage to capture critical characteristics of contact states during switching operations. The software layer proposes a dual-stage optimization strategy: In the data preprocessing stage, an image enhancement algorithm combining geometric transformations (rotation/cropping), histogram equalization, image filtering, and noise injection is developed to significantly improve image quality and augment sample diversity. During the feature extraction phase, a deep-optimized convolutional neural network(CNN) model incorporating an efficient channel attention (ECA)mechanism was designed. The efficient channel attention mechanism module enables dynamic allocation of channel feature weights, thereby enhancing the extraction of key features. Algorithm analysis and verification show that the designed image enhancement algorithm can effectively improve image quality, highlight key features, and increase the accuracy of the recognition model by over 30%. The proposed ECA-CNN model can further enhance the focus on key features of the image on the basis of the image enhancement algorithm, with a recognition accuracy rate of over 97%, which is 12% higher than that of the CNN model.
Key words: high-voltage disconnect switch; state recognition; image enhancement; convolutional neural network; attention mechanism
參考文獻(xiàn)
[1] 蔣健,陳宇昇,蔡潤慶,等 . 電網(wǎng)高壓隔離開關(guān)機(jī)械故障診斷 [J]. 沈陽工業(yè)大學(xué)學(xué)報(bào),2023,45(2):127-132.
[2] 劉偉杰,趙軍亮 . 一起隔離開關(guān)傳動機(jī)構(gòu)失效原因分析及改進(jìn) [J]. 高壓電器,2024,60(8):269-274.
[3] 公多虎,詹仲強(qiáng),葛志杰,等 . 基于小波分析的高壓隔離開關(guān)故障檢測方法研究 [J]. 自動化技術(shù)與應(yīng)用, 2024,43(3):19-22.
[4] 關(guān)為民,卞超,李東風(fēng) . 基于主軸扭矩-轉(zhuǎn)角特性 的高壓隔離開關(guān)缺陷分析仿真 [J]. 高壓電器,2024, 60(1):48-54.
[5] 王旭鵬,郝建,鐘堯,等 . 基于 ILPSO-LSSVM的GIS設(shè)備隔離開關(guān)機(jī)械異響振動識別方法研究 [J]. 高壓電器,2023,59(11):172-183.
[6] 顧曉東,唐丹宏,黃曉華 . 基于深度學(xué)習(xí)的電網(wǎng)巡檢圖像缺陷檢測與識別 [J]. 電力系統(tǒng)保護(hù)與控制, 2021,49(5) :91-97.
[7] 袁偉,趙玉凱,李楊,等 . 應(yīng)用于光纖的高壓隔離開關(guān)狀態(tài)識別技術(shù) [J]. 傳感器與微系統(tǒng),2024, 43(8):160-164.
[8] 杜耀文,謝靜,劉志堅(jiān),等 . 基于深度學(xué)習(xí)的高壓隔離開關(guān)分合狀態(tài)檢測算法研究 [J]. 電力系統(tǒng)保護(hù)與控 制,2023,51(19) :114-123.
[9] 劉子英,張靖,鄧芳明 . 基于 B P 神經(jīng)網(wǎng)絡(luò)的高壓隔離開關(guān)分合閘監(jiān)測識別 [J]. 電力系統(tǒng)保護(hù)與控制, 2020,48(5):134-140.
[10] 張靖,單長吉,周麗,等 . 基于圖像融合的高壓隔離開關(guān)分合閘狀態(tài)識別 [J]. 紅外技術(shù),2024,46(5):539-547.
[11] 陳富國,蔡杰,李中旗 . 基于機(jī)器視覺的高壓隔離開關(guān)設(shè)備狀態(tài)判別與故障診斷技術(shù) [J]. 微型電腦應(yīng)用, 2022,38(2):191-194.
[12] 萬嘉龍,況立群,曹亞明,等 . 復(fù)合主干融合的圖像增強(qiáng)算法 [J]. 計(jì)算機(jī)系統(tǒng)應(yīng)用,2024,33(4):194-201.
[13] WANG Qilong, WU Banggu, ZHU Pengfei, et al.ECANet: Efficient channel attention for deep convolutional neural networks[C]//2020 IEEE/ CVF Conference on Computer Vision and Pattern Recognition,2020 :11534-11542.
[14] 馬群,趙美蓉,鄭葉龍,等 . 基于自適應(yīng)條件直方圖均衡的紅外圖像細(xì)節(jié)增強(qiáng)算法 [J]. 紅外技術(shù),2024, 46(1):52-60.
[15] 石野,王曉凱,劉慧鋒 . 基于局部濾波模板的雙邊濾波算法研究 [J]. 測試技術(shù)學(xué)報(bào),2021,35(1):49-53.