There are problems of poor accuracy and low efficiency caused by the lack of data in traditional three-dimensional reconstruction methods. Thus, based on a three-dimensional laser scanning point cloud, a recognition method for substation equipment is proposed. This combines the particle swarm optimization algorithm and the k-nearest neighbor classification algorithm. The particle population optimization algorithm is used to optimize the coefficient weight of each subspace feature, and the k-nearest neighbor classification method is used to classify the equipment. The influence of the size and loss rate of the point cloud subspace on the recognition effect is analyzed through experiment. It is compared with the improved iterative closest point algorithm to verify the superiority and accuracy of the method. Experimental results show that this method can effectively improve recognition accuracy and efficiency. The recognition accuracy can reach more than 95%, and the average recognition time is 0.19 seconds, which has application value. This work is supported by the Science and Technology Project of State Grid Corporation of China (No. 5216A0182020R).