【摘要】 结合粗糙集理论,提出一种传感器数学模型客观确定方法.以移动机器人本体上应用最广泛的声纳传感器为例,分析声纳采集的大量客观数据之间的依赖关系,利用粗糙集理论的分类思想中正域、依赖度等概念,对按照一定规则离散化的声纳数据进行计算,获取传感器各个离散点的取样值,存入数据库中以备需要时查询.以PioneerⅡ机器人为实验平台,验证了方法的正确性和有效性.该方法可以推广到其他类型传感器建模中,不需要大量各领域专家的经验和知识,仅仅需要足够的采集数据即可建立不同的传感器模型,既避免了传感器建模的主观性又具有一定的通用性.更多还原

【Abstract】 A new mathematical model of sensor was obtained in combination with rough sets(RS) theory.Sonar sensors which were most widely used in the mobile robot body were introduced as an example.In this paper,by analyzing of the dependency degree among the attributes based on RS,a value to express mathematical modeling was given.The classification method and concepts about positive domain and dependence of RS were introduced only depended on the massive objective data collected by sonar sensors.At last,Pioneer II mobile robot was used and the correctness and validity of the developed method were verified with a particular instance.The method can be extended to other types of sensors modeling.It thus avoids that the different sensor models need massive expertise and knowledge of their respective areas.Only sufficient data will be needed.Not only the subjective during the sensor modeling is avoided but also the universal is keeping.更多还原