面向无人机电力巡检航迹规划的语义服务方法

Semantic method for trajectory planning of UAV power line inspection

针对复杂环境下无人机电力巡检航迹规划问题,从信息处理角度提出一种语义服务方法。构建一个具有环境态势感知、航迹构造和语义策略图模型的语义服务框架;利用巡检关联比重对实体对象进行空间划分,通过OWL形式化表示和Jena推理刻画巡检航线特征位置的因果关系,提出语义空间距离计算和基于强化学习的路径选择机制。研发语义服务模拟演示系统,实验结果表明,在满足航向速率和空速的条件下,通过语义策略图和强化学习方法可以有效判断飞行特征点及干扰因素,同时语义规划的航迹代价能够快速收敛并趋于稳定,巡检对象覆盖率达到92.3%,实现无人机航迹规划在电力巡检的应用。

Focusing on the issue of trajectory planning for unmanned aerial vehicle(UAV)power inspection in complex environment,a semantic service method was proposed from the perspective of information processing.A semantic service framework including environment awareness,trajectory construction and semantic-strategy map model was constructed.The spatial for entity objects was divided which was made use of inspection related proportion,the causal relationship of patrol route location nodes was depicted by owl formal representation and Jena reasoning,and an algorithm of semantic spatial distance and route path selection mechanism based on reinforcement learning was proposed.A demonstration system of semantic service simulation was carried out.Experimental results show that the proposed method can judge the flight feature points and interference factors through semantic-strategy map and reinforcement learning in the case of satisfying rate and speed conditions,while the trajectory cost can converge to the stable value rapidly and the coverage rate of inspection object can cover more than 90%,which realizes the application of UAV trajectory planning in power inspection.