Indoor intelligent logistics robots are widely employed in warehousing and sorting operations, where their autonomous obstacle avoidance capabilities play a pivotal role in enhancing both operational efficiency and system safety. To bolster the obstacle avoidance reliability of individual robots and improve the collaborative path planning efficiency within multi-robot systems, comprehensive research is currently being conducted. This research primarily focuses on acquiring environmental information by integrating optical flow and feature extraction techniques. Furthermore, it involves the fusion of multi-sensor data from depth perception cameras and laser radars to enhance the precision of environmental perception. Additionally, the Pelican optimization algorithm has been refined through the incorporation of chaotic mapping and firefly disturbance strategies, with the overarching goal of achieving optimal path planning. The experimental setup includes obstacle avoidance simulation testing on the Ubuntu 20.04 platform, along with on-site testing at an e-commerce sorting center. The experimental results demonstrate that, compared to traditional methods, the improved approach increases the average obstacle avoidance safety distance by over 60% in static environments and enables flexible responses to random obstacles, such as pedestrians, in dynamic settings. The refined Pelican optimization algorithm exhibits superior convergence speed and accuracy in both single-peak and multi-peak test functions. In real-world logistics scenarios, the obstacle avoidance success rate reaches 98.6%, with an average obstacle avoidance time of 1.5 seconds and a total path length of 12.5 meters—all of which outperform traditional methods and existing technologies. The visual obstacle avoidance technology proposed in this study effectively enhances the obstacle avoidance safety of robots and the efficiency of multi-robot collaborative operations, offering a reliable solution for indoor logistics robots to autonomously navigate and avoid obstacles.
Jiaying Li (Tue,) studied this question.