论文部分内容阅读
针对移动机器人上的声纳传感器感知周围环境具有很大的不确定性和不精确性问题,提出DSmT融合机技术(包括信息过滤器、融合算子和冲突分配器),在建立声纳信度赋值模型的基础上,首先利用基于证据支持贴近度函数的信息过滤器过滤掉部分不一致信息,然后利用经典DSm组合规则和PCR5冲突分配规则,融合16个声纳的不确定感知信息,使移动机器人能够准确地感知非结构环境.最后通过让机器人在一个相对拥挤具有家居特征的实验室环境中运行,得到了仅次于激光传感器感知环境的效果,充分地表明了该方法的有效性.
Aimed at the great uncertainty and inaccuracy of the sensing environment of sonar sensors on mobile robots, DSmT fusion machine technology (including information filter, fusion operator and conflict allocator) is proposed. After establishing the sonar reliability Based on the assignment model, we first filter some of the inconsistent information by using the information filter based on the evidence support proximity function, and then use the classic DSm combination rule and the PCR5 conflict assignment rule to fuse the 16 sonar uncertainty perception information, and make the mobile robot Which can accurately sense the unstructured environment.Finally, by letting the robot run in a relatively crowded and home-like laboratory environment, the effect of the environment after the laser sensor is obtained, which fully demonstrates the effectiveness of the method.