关键词: 变电站机器人, 时空网络, 注意力结构, 分类回归预测, 相似特征过滤

Abstract:

In order to improve the visual obstacle avoidance ability of substation robots in complex environments, a robot visual obstacle avoidance method based on spatiotemporal networks is proposed. The method utilizes traditional image processing techniques to enhance road information and designs a lightweight deep convolutional neural network structure to extract road features from a spatial domain perspective; based on the spatial characteristics of the road, a long short-term memory network is introduced to mine the changes in the road from a temporal perspective, and a classification regression prediction structure is used to predict the robot's obstacle avoidance direction and angle; considering the highly repetitive characteristics of substation roads, a feature filtering module is introduced to reduce redundant feature calculations and ensure real-time performance during network applications. The experimental results show that the proposed method can effectively extract spatiotemporal features of substation road scenes, this method can more accurately predict the next action of the robot and better complete navigation obstacle avoidance tasks.

Key words: substation robot, spatiotemporal network, attention structure, classification regression prediction, similar feature filtering