摘要
针对船载传感器信息的多源性、多维性和异构性特征,构建船舶动态信息融合模型,以提高内河航行中船员对目标船动态的感知和辨析能力。通过分析船载导航设备在船舶动态辨析中的局限性和互补性,构建集成卡尔曼滤波、自适应加权和神经网络等方法的多源异构数据融合模型,并通过实测和模拟数据计算对该模型进行验证,确定系统的可靠性、稳定性和精确性。该模型不仅对保障内河船舶航行安全、提高内河水上交通效率具有重要的理论参考价值,而且对内河水运安全的保障工作具有积极的现实意义。
A dynamic information fusion model specifically for multi-source,multi-dimensional and heterogeneous information of shipborne sensors is constructed to improve the detection and discrimination of target ships in inland waterways. This Kalman filter based model for heterogeneous data integration,featuring the adaptive weight fusion method and neural network function,is constructed based on the analysis of the limitations and complementarities of different shipborne navigation equipment for identifying target ships. The reliability,stability and accuracy of the model are checked through comparing simulation output to the measured data. The information fusion model can be useful means for inland water transport safety.
出处
《中国航海》
CSCD
北大核心
2017年第4期16-20,共5页
Navigation of China
基金
国家自然科学基金面上项目(51179147)
