摘要
针对AUV在高度不确定海洋环境中执行任务的过程中需要实时准确地感知出当前环境、系统状态和任务执行中不确定事件对任务成功的影响,并为AUV任务重规划提供触发条件的问题,提出了一个具有不确定事件检测、不确定事件识别和不确定环境本体模型的环境感知框架,实现了基于本体推理和模糊逻辑结合的不确定事件检测方法和基于贝叶斯网络的不确定事件识别方法,并利用不确定事件检测和不确定事件识别结果更新具有概率扩展的不确定环境本体模型,提高本体在不确定知识的表示和推理方面的能力.结合AUV在不确定海洋环境中执行导航任务的应用背景,对AUV环境感知进行仿真实验并分析了实验结果,验证了不确定海洋环境下AUV环境感知方法的有效性.
It is necessary for AUV to real time perceive the influence degree of current uncertain events of ocean environment, AUV status and mission executive on the task, and to provide trigger condition to re-planning during mission. An environment perception for AUV in uncertain ocean environment is proposed. Environment perception makes use of the idea that heterogeneous real-world data of different types must be processed by uncertain event detection layer and uncertain event recognition layer. The uncertain event detection method is based on ontology reasoning and fuzzy logic, and the uncertain event recognition method is based on Bayesian network . The results of uncertain event detection and uncertain event recognition are used to the update uncertain environment ontology mode extended by probability description. In this way, environment perception method for AUV in uncertain ocean environment extends the ability of uncertainty presentation and reasoning in ontology. Navigation experimental results show the effectiveness of environment perception method for AUV in uncertain ocean environment.
出处
《计算机研究与发展》
EI
CSCD
北大核心
2013年第9期1981-1991,共11页
Journal of Computer Research and Development
基金
国家自然科学基金项目(60975071
61100005)
关键词
智能水下机器人
不确定海洋环境
环境感知
不确定事件检测
不确定事件识别
环境本体模型
autonomous underwater vehicle (AUV)
uncertain ocean environment
environmentperception
uncertain event detection
uncertain event recognition
environment ontology model