机器人
后验概率
职位(财务)
声发射
计算机科学
贝叶斯概率
碰撞
贝叶斯推理
干扰(通信)
多向性
概率逻辑
点(几何)
算法
到达时间
点源
声学
物理
数学
人工智能
光学
几何学
电信
频道(广播)
计算机安全
财务
节点(物理)
经济
无线
作者
Zihui Chen,Zhinong Li,Fengshou Gu
标识
DOI:10.1088/1361-6501/ad093c
摘要
Abstract The existing source localization based on acoustic emission technology often depends on the assumption of constant wave velocity inside the material. However, this assumption is hardly satisfied in actual engineering. The uncertainty of wave velocity can easily lead to low localization accuracy of sweeping robots. To overcome these deficiencies, a complete probability multi-directional measurement method based on the Bayesian inference mechanism is proposed. In the proposed method, based on the Bayesian probabilistic model, the extracted sensor time difference is subjected to probabilistic inference using the coordinate input model to determine the posterior distribution of the source’s position and wave velocity of the given arrival time. Compared with the traditional time-difference method, the proposed method achieves excellent results and outperforms the standard time-difference method in localization accuracy and anti-interference. In addition, the proposed method can conveniently, quickly, and effectively determine the location of the colliding point without considering the source emission time and wave velocity. The research in this paper provides an effective method for solving the collision localization problem of the sweeping robot shell under the acoustic emission time and wave velocity are unknown.
科研通智能强力驱动
Strongly Powered by AbleSci AI