基于模糊距离和DS证据融合的清洁机器人定位设计

针对现有清洁机器人定位算法具有的定位精度不高且难以保证在线定位的实时性问题,设计了一种基于模糊距离和DS(DEMPSTER-SHAFER)证据融合的清洁机器人自定位算法;首先,对清洁机器人的总体结构进行了建模和分析,并对其进行硬件设计,为了实现清洁机器人的实时精确定位,采用DS证据数据融合方法对多传感器采集的数据进行有效数据融合,为了进一步提高其精确性,引入了模糊距离,定义当前传感器采集数据与理想结果之间的距离,根据距模糊距离的大小自适应地调大或减少传感器采集数据分配信度的权重,将加权信度作为新的信度进行融合得到最终的融合结果;在不同的场地中进行实际试验,对文中设计的清洁机器人进行实际定位,实验结果表明文中方法能有效地进行定位,较经典的DS证据融合和其它方法具有较高的定位精度,且具有较小的时间复杂度和空间复杂度,具有较大的优越性。

Localization Design for Cleaning Robot Based on Fuzzy distance and DS Evidence Fusion

Aiming at the given cleaning robot having the defects of low localization accuracy and not has the ability of in time localizing with the on-line circumstance, a localization algorithm based on fuzzy distance and DS(DEMPSTER-SHAFER) evidence fusion is proposed. Firstly, the main structure of cleaning robot is modeled and analyzed, and the hardware is designed. In order to realize its accurate localization, the fuzzy distance is introduced, the distance between the data collected by current sensor node and the ideal result, according the value of the distance to adaptively and dynamically adjust the weight for the basis belief allocation weight, the weighted belief is used as the new belief to get the final fusion result. The experiment is operated in several fields and the designed cleaning robot is traveling in the fields to collect the rubbish and self-localizing, the result shows the method in this paper can realize localization for cleaning robot, compared with the DS evidence fusion method and other methods, it has the higher localization precision with less time and space complexity, therefore, it has big priority.