为实现智能避障，确保船舶航行安全，设计基于网络感知信息的船舶航迹规划智能避障方法。高效感知船舶本身及附近环境信息，精准获取船舶本身及附近船舶位置信息。依据位置信息构建船舶航迹规划中碰撞危险度模型，判断本船是否存在碰撞危险。若存在碰撞危险，则采用模糊神经网络智能控制船舶航行方向完成智能避障。试验结果表明：该方法可精准感附近信息，在简单与复杂航迹环境下均能够实现智能避障；本船与附近船舶距离高于安全距离，船航向角变化曲线平稳、无抖动情况，确保船舶航行安全。
In order to realize intelligent obstacle avoidance and ensure the safety of ship navigation, an intelligent obstacle avoidance method for ship track planning based on network sensing information is designed. Efficiently perceive the information of the ship itself and its surrounding environment, and accurately obtain the location information of the ship itself and its surrounding ships; According to the position information, the collision risk model in ship track planning is constructed to judge whether the ship has collision risk; If there is a collision risk, the fuzzy neural network is used to intelligently control the ship navigation direction to complete the intelligent obstacle avoidance. Experiments show that this method can accurately sense the nearby information and realize intelligent obstacle avoidance in both simple and complex track environments; The distance between the ship and nearby ships is higher than the safe distance, and the change curve of ship heading angle is stable without jitter, so as to ensure the safety of ship navigation.
2022,44(3): 71-74 收稿日期：2021-08-23
DOI：10.3404/j.issn.1672-7649.2022.03.014
分类号：U675.96
基金项目：宁夏自治区级教改项目(NXJG2017087)
作者简介：黄磊(1981-)，男，硕士，副教授，主要研究领域为计算机网络
参考文献：
[1] 魏万淇, 高曙, 初秀民. 基于流推理的定线制水域船舶航行安全评估方法[J]. 中国安全科学学报, 2019, 29(3): 168–173
[2] 罗贤程, 尹建川, 李宏宾. 基于动态分阶势场法的船舶自动避碰系统[J]. 上海海事大学学报, 2020, 41(1): 1–7
[3] 刘忠, 伊戈, 张建强. 基于改进人工势场法的无人艇避障算法[J]. 海军工程大学学报, 2021, 33(5): 28–32
[4] 王百勇, 张艳华, 贾俊乾. 基于深度学习理论下电子海图与雷达图像船舶感知信息融合[J]. 现代雷达, 2021, 43(5): 44–50
[5] 肖长诗, 陈芊芊, 文元桥, 等. 基于LoRa的封闭水域船舶自动识别系统设计[J]. 船海工程, 2019, 48(1): 158–162+167
[6] 秦操, 杨荣武. 复杂水域慎思型避碰轨迹规划[J]. 上海海事大学学报, 2020, 41(3): 12–18
