船舶驾驶台值班监测系统关键技术研究

发布时间：2018-03-03 09:14

本文选题：船舶驾驶　切入点：值班监测　出处：《集美大学》2017年硕士论文　论文类型：学位论文

【摘要】：船舶航行安全一直都是海事领域研究的焦点问题。据相关的统计资料显示,所发生的海上船舶安全事故中有80%以上是由于人为因素造成的,而这些由人为因素导致的海上事故中50%以上是由于船员疲劳导致的。疲劳使船员控制力和注意力下降,避碰反应速度慢,船舶操纵质量下降,最终导致海上事故。为了监测值班船员的警觉性,国际海事组织(IMO)通过决议MSC.128(75),强制船舶安装船桥航行值班报警系统(BNWAS,Bridge Navigational Watch Alarm System)。系统需要船员在规定时间内手动复位,该方式增加船员的工作负担,不利于船舶航行安全。针对BNWAS存在的不足,本文提出一种基于人体姿势和步态的疲劳检测方法。因为人体行为能力受神经控制,以及骨骼肌肉的支撑,当人的精神饱满时,肌肉力量充足,行为能力正常,但人体疲劳时人体行为能力会有所下降。本文提出了基于向量角度的姿势识别算法和使用改进的DTW算法识别异常步态。研究分析传感器的硬件结构、工作原理、软件开发工具包。应用OpenCV视觉库对获取的图像进行处理和显示。使用C++、OpenCV、Kinect SDK开发船舶驾驶台值班船员监测系统,并通过实验进行验证算法的有效性。
[Abstract]:Ship navigation safety has always been the focus of maritime research. According to the relevant statistical data, more than 80% of the maritime ship safety accidents are caused by human factors. More than 50% of these accidents at sea caused by human factors are caused by crew fatigue. Fatigue reduces crew control and attention, slow collision avoidance reaction, and lower ship handling quality. In order to monitor the alertness of the crew on duty, the International Maritime Organization (IMO) adopted resolution MSC.12875 to force ships to install the ship bridge navigation watch alarm system (BNWASA Bridge Navigational Watch Alarm system). The system requires the crew to manually reset within a specified time. This method increases the work burden of the crew and is not conducive to the safety of ship navigation. In view of the shortcomings of BNWAS, a fatigue detection method based on human posture and gait is proposed in this paper. And the support of skeletal muscles, when people are full of energy, they have plenty of muscle strength and normal behavior. However, the ability of human behavior will decrease when human body is tired. In this paper, a pose recognition algorithm based on vector angle and an improved DTW algorithm are proposed to identify abnormal gait. The hardware structure and working principle of the sensor are studied and analyzed. Software development toolkits. The OpenCV visual library is used to process and display the obtained images, and C # OpenCVCV Kinect SDK is used to develop the ship bridge crew watch monitoring system, and the validity of the algorithm is verified by experiments.
【学位授予单位】：集美大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U698

【参考文献】