针对移动机器人基于超声波传感器的地图创建,提出了一种双重环境模型。该模型同时包涵以直线段表示的几何模型和以概率形式表示的栅格模型,结合了几何模型与栅格模型的优点。给出了该模型的建立及更新算法,并基于该模型提出了一种主动探测策略,在保证探测效率的同时减小了超声波传感器的镜面反射产生的不良影响。在室内环境下进行移动机器人实验,验证了基于双重环境模型的探测策略的有效性。
A dual-representation environment model is presented for the map building problem using ultrasonic sensors. The model contains the geometric model and the inference grid model combining the advantages for both models. The algorithm for founding and updating the model is presented. Based on the model, an active exploration strategy is proposed to reduce the effect of the specular reflection while ensuring the efficiency of map building. The experiment is executed in the indoor environment to verify the performance of the algorithm.