An Improved Substation Equipment Recognition Algorithm by KNN Classification of Subspace Feature Vector
Liang Wang, Qilong Kou, Rongle Wei, Leiyue Zhou, Tao Fang, Jia Zhang · 2021 China Automation Congress (CAC) · 2021
With the development of intelligent substation construction technology, 3D recognition of substation equipment is becoming more and more important. This paper proposes an improved substation equipment recognition algorithm based on KNN classification of subspace feature vector, in which the number of dividing subspaces is unchanged to maximize keeping the shape features of equipment point cloud, and solves the problem of distorting shape feature of point cloud caused by normalization. At the same time, this paper preselects the standard devices from the template library by size screening of standard devices to reduce the comparison range of recognition and shorten average recognition time. The experiment results show that recognition accuracy of the improved recognition algorithm is improved by 2%-8% and the average recognition time is reduced by an order of magnitude, which shows the effectiveness of the recognition method of this paper.