Working environment of mobile robots has gradually expanded from indoor structured scenes to outdoor scenes such as wild areas in recent years. The expansion of application scene, change of sensors and the diversity of working tasks bring greater challenges and higher demands to active localization for mobile robots. The efficiency and stability of traditional localization strategies in wild environments are significantly reduced. On the basis of considering features of the environment and the robot motion curved surface, this paper proposes a curve-localizability-SVM active localization algorithm. Firstly, we present a curve-localizability-index based on 3D observation model, and then based on this index, a curve-localizability-SVM path planning strategy and an improved active localization method are proposed. Obtained by setting the constraint space and objective function of the planning algorithm, where curve-localizability is the main constraint, the path helps improve the convergence speed and stability in complex environments of the active localization algorithm. Helped by SVM, the path is smoother and safer for large robots. The algorithm was tested by comparative experiments and analysis in real environment and robot platform, which verified the improvement of efficiency and stability of the new strategy.

中文翻译：

---

户外环境中移动机器人的曲线可定位性-SVM主动定位研究

近年来，移动机器人的工作环境已从室内结构化场景逐渐扩展到野外等室外场景。应用场景的扩展，传感器的更换以及工作任务的多样性给移动机器人的主动定位带来了更大的挑战和更高的要求。传统定位策略在野生环境中的效率和稳定性大大降低。在考虑环境特征和机器人运动曲面的​​基础上，提出了一种曲线可定位性-SVM主动定位算法。首先，提出了一种基于3D观测模型的曲线可定位性指标，然后基于该指标提出了一种曲线可定位性SVM路径规划策略和一种改进的主动定位方法。通过设置规划算法的约束空间和目标函数（以曲线可定位性为主要约束）来获得路径，这有助于提高主动定位算法在复杂环境中的收敛速度和稳定性。在SVM的帮助下，大型机器人的路径更平滑，更安全。通过在真实环境和机器人平台上的对比实验和分析对算法进行了测试，验证了该策略的效率和稳定性的提高。