基于机器视觉的图书机器人取书路径控制方法研究

Research on book robot book fetching path control method based on machine vision

摘要:

在图书馆自动化和智能化发展的背景下,实现高效的图书机器人取书路径规划成为一项关键的技术挑战。鉴于此,研究尝试采用利用机器视觉的图书机器人取书路径控制方法。首先,研究利用实时定位与地图构建技术中的Cartographer算法结合激光视觉雷达构建了图书馆栅格电子地图。其次,引入A*算法并通过对优化后的蚁群优化算法进行结合,实现并提出了一种新型机器人取书路线规划控制模型,经实验测试发现,当权重系数为0.6时,新型算法模型的性能表现最佳。其规划的方案路径最短为17m,成功率最高为95%。仿真测试后发现,新模型的线路规划拐点数最少为10个,相较于同类型模型减少了4个拐点。由此可知,研究所提模型在路径长度和成功率方面均优于传统算法,有效提升了图书机器人的取书效率和准确性,能够为机器视觉与移动机器人技术的应用研究贡献一种新的思路。
关键词:机器视觉；A*算法；电子地图；线路规划；取书
中图分类号:TP242.2 ????? 文献标识码:A

Abstract:

In the context of library automation and intelligent development, achieving efficient book robot book fetching path planning has become a key technical challenge. In view of this, the study tries to adopt a book robot book fetching path control method using machine vision. First, the study utilizes the Cartographer algorithm in real-time localization and map construction technology combined with laser vision radar to construct a library raster electronic map. Secondly, A* algorithm is introduced and by combining the optimized ant colony optimization algorithm, a new robot book fetching route planning and control model is implemented and proposed, and it is found by experimental testing that the new algorithm model has the best performance when the weight coefficient is 0.6. The shortest path of its planned program is 17m, and the highest success rate is 95%. After simulation test, it is found that the new model has a minimum number of 10 inflection points for line planning, which is 4 inflection points less compared to the same type of model. It can be seen that the proposed model of the study is better than the traditional algorithm in terms of path length and success rate, which effectively improves the efficiency and accuracy of book fetching by the book robot, and can contribute a new way of thinking to the applied research of machine vision and mobile robot technology.