摘要
异常行为的出现,会对船舶正常航行造成较大影响。为避免上述现象的出现,设计基于模式识别的船舶异常行为自动报警方法。通过主成分分析、成分因子相关性分析2个步骤,完成基于模式识别的船舶异常行为特征提取。通过异常行为的局部决策、基于决策结果的数据融合2个步骤,完成基于模式识别船舶异常行为自动报警方法的搭建。模拟方法运行环境,设计对比实验结果表明,应用基于模式识别船舶异常行为自动报警方法,可以明显降低船舶异常行为的发生几率,为船舶正常航行提供有力保障。
The occurrence of abnormal behavior will have a great influence on the normal navigation of ships. In order to avoid these phenomena, we design an automatic alarm method for ship abnormal behavior based on pattern recognition.Through two steps of principal component analysis and component factor correlation analysis, we complete the feature extraction of ship abnormal behavior based on pattern recognition. By the two steps of local decision of abnormal behavior and data fusion based on decision results, the method of automatic alarm for ship abnormal behavior based on pattern recognition is built. The simulation results show that the automatic alarm method based on pattern recognition for ship abnormal behavior can obviously reduce the probability of the abnormal behavior of the ship and provide a powerful guarantee for the normal navigation of the ship.
出处
《舰船科学技术》
北大核心
2018年第7X期43-45,共3页
Ship Science and Technology
关键词
模式识别
船舶异常行为
自动报警
主成分分析
相关性分析
局部决策
数据融合
pattern recognition
ship abnormal behavior
automatic alarm
principal component analysis
correlation analysis
local decision making
data fusion
