【摘要】：針對我國高壓輸電線路巡檢周期長、作業(yè)強(qiáng)度大、無法保證巡檢結(jié)果客觀性與完整性的問題,提出了一套無人機(jī)三維激光雷達(dá)智能識別輸電線路缺陷的研究方法。該方法首先使用自動化分類方法將走廊三維點云進(jìn)行金字塔式數(shù)據(jù)重采樣及分檔式處理;其次將分檔數(shù)據(jù)自動劃分為地面、植被、桿塔、導(dǎo)線4個類別,結(jié)合人工判讀識別低等電力線、公路等其他類別;最終,根據(jù)國網(wǎng)相關(guān)安全局規(guī)程提取當(dāng)前工況下導(dǎo)線與周圍地物存在的安全缺陷。實驗數(shù)據(jù)表明,所提出的方法可有效識別導(dǎo)線-植被、導(dǎo)線-建筑物、導(dǎo)線-交跨等多類重要缺陷,可大幅提升現(xiàn)有高壓輸電線路巡檢流程的質(zhì)量與效率。
[Abstract]:Aiming at the problem that the inspection period of HV transmission lines in China is long and the intensity of operation is too large to guarantee the objectivity and integrity of the inspection results, a set of intelligent identification methods for 3D lidar of UAV to identify the defects of transmission lines is proposed. The method firstly uses automatic classification method to carry out pyramid data resampling and file processing of corridor 3D point cloud. Secondly, the classified data are automatically divided into four categories: ground, vegetation, tower and wire, combined with manual interpretation to identify the lower power line, highway and other categories; Finally, according to the relevant safety rules of the national network, the safety defects of the wire and the surrounding objects are extracted under the current operating conditions. Experimental data show that the proposed method can effectively identify many kinds of important defects such as traverse vegetation traverse-building traverse-intersection and so on and can greatly improve the quality and efficiency of the existing high voltage transmission line inspection process.
【作者單位】： 北京拓維思科技有限公司;南方電網(wǎng)科學(xué)科研院有限責(zé)任公司;天津航天中為數(shù)據(jù)系統(tǒng)科技有限公司;天津市智能遙感信息處理技術(shù)企業(yè)重點實驗室;云南中能投資股份有限公司楚雄直升機(jī)作業(yè)分公司;
【分類號】：TM75;TN958.98
，

本文編號：2418105