中文摘要：

高压设备运行过程中会发生电晕放电现象,其危害很大。本文利用紫外放电检测新技术,融合采集到电晕和可见光图像,准确对故障点进行定位。采用遗传算法的投影寻踪理论,利用Delphi软件对采集的紫外电晕图进行处理后作为分级数据,建立遗传投影寻踪等级模型,对电晕放电强度的量化分级研究。实验结果验证了该模型的合理性,测试结果也达到预期的结果,对放电故障检测有一定的实际应用意义。

英文摘要：

High voltage equipment that runs for a long time could produce corona discharge phenomenon. The harm of corona discharge is serious, so it＇s necessary to detect the corona. As a new technology of detecting high pressure discharge equipment fault, this article used ultraviolet discharge detection that can fuse to corona and visible light images and can locate fault point accurately. The article used the theory of projection pursuit in genetic algorithm, and used the collection of ultraviolet corona figure after processing of the Delphi software as a hierarchical data, and the level model of genetic projection pursuit was set up to study the quantitative classification of corona discharge strength. The experimental results verified the rationality of the model. Test results achieve the desired requirements, and have certain practical significance to discharge fault detection.