为了改进与补充当前船舶机舱的操作指导，提出一种基于贝叶斯网络推理的操作推荐方法。首先，利用收集和分析实际机舱操作间逻辑关系与模拟器历史操作数据构建出贝叶斯网络模型；其次，结合模拟器操作的历史数据和专家经验确定设备状态与操作节点的先验概率表和条件概率表；最后，根据准确率优化贝叶斯推理模型并完成操作推荐。经过实验对比，可以得到基于贝叶斯网络的船舶机舱操作指引方法在应急发电机启动过程中42个节点操作指导准确率为92.86%，方差为0.0008。在基于贝叶斯决策推理的船舶机舱操作推荐方法帮助下，能够有效确保操作的准确性和针对不同情况推荐的多样性。
In order to improve and supplement the current operation guidance of ship engine room, an operation recommendation method based on Bayesian network inference was proposed. Firstly, the Bayesian network model was constructed by collecting and analyzing the logical relationship between the actual cabin operations and the historical operation data of the simulator. Secondly, combined with the historical data and expert experience of the simulator operation, the prior probability table and the conditional probability table of the equipment state and the operation node were determined. Finally, the Bayesian inference model was optimized according to the accuracy and the operation recommendation was completed. After experimental comparison, it can be obtained that the operation guidance method of ship engine room based on Bayesian network has an accuracy of 92.86% and a variance of 0.0008 for 42 nodes in the process of emergency generator starting. With the help of the ship engine room operation recommendation method based on Bayesian decision reasoning, the accuracy of operation and the diversity of recommendations for different situations can be effectively ensured.
2025,47(22): 53-60 收稿日期：2024-11-5
DOI：10.3404/j.issn.1672-7649.2025.22.008
分类号：U664.82+1
基金项目：国家重点研发计划资助项目（2022YFB4301400）；高技术船舶科研项目（CBG3N21-3-3）
作者简介：宋志豪（1999 – ），男，硕士研究生，研究方向为轮机自动化与智能化
参考文献：
[1] 钱兴达. 船舶机舱温度检测采集及预测关键技术研究[D]. 镇江: 江苏科技大学, 2018.
[2] 贾宝柱, 林叶锦, 曹辉, 等. 轮机模拟器中机舱资源管理培训及评估功能[J]. 中国航海, 2013, 36(3): 28-33.
[3] 李军, 朱云升. 三角模糊层次分析的综合评价在黄河库区船舶安全评估中的应用[J]. 重庆交通大学学报(自然科学版), 2021, 40(7): 16-22.
[4] 许萌萌, 张成伟, 梅顺峰, 等. 基于数据大脑的船岸一体机舱智能运维系统研究设计[J]. 中国舰船研究, 2022, 17(6): 79-87.
[5] 伍巧芸. 船舶轮机模拟器的应用与发展[J]. 中国水运(下半月), 2023, 23(5): 43-45.
[6] 郎舒妍, 曾晓光, 张民. 智能船舶工程科技发展战略研究[J]. 中国工程科学, 2019, 21(6): 27-32.
[7] 曹辉, 梁世源, 赵辉, 等. 基于TOPSIS-Vague集的轮机模拟器智能评估方法[J]. 重庆交通大学学报(自然科学版), 2022, 41(11): 1-6+24.
[8] LOPEZ M A A, FLORES C H, GARCIA E G. An intelligent tutoring system for turbine startup training of electrical power plant operators[J]. Expert Systems with Applications, 2003, 24(1): 95-101.
[9] HASHIMOTO Y, KITAMURAL Y, OHASHI T, et al. Transient model-based operation guidance on blast furnace[J]. Control Engineering Practice, 2019, 82: 130-141.
[10] STEDANO R. The use of Petri Nets models in railway traffic applications[J]. IFAC Proceedings Volumes, 2009, 42(5): 151-156.
[11] 李孝敏. 船舶轮机设备故障处理分析[J]. 智能城市应用, 2023, 6(7): 73-75.
[12] 王源庆, 苏娟. 虚拟现实轮机模拟器在航运人才培训中的应用[J]. 中国航务周刊, 2024(33): 51-53.
[13] 曾青山, 李忠辉. WMS2000型轮机模拟器在航海教学中的应用与展望[J]. 航海教育研究, 2002, 19(4): 61-63.
[14] 曹威. 高效的贝叶斯推理电路设计及应用[D]. 成都: 电子科技大学, 2019.
[15] 胡勇健, 肖志怀, 周云飞, 等. 基于贝叶斯网络Noisy Or模型的水电机组故障诊断研究[J]. 水力发电学报, 2015, 34(6): 197-203.
[16] 刘亚辉, 谭暑秋, 王越. 一种改进的贝叶斯网络结构学习分类模型及应用[J]. 计算机应用与软件, 2013, 30(3): 118-121+144.
[17] COOPER G F, HERSKOVITS E. A Bayesian method for the induction of probabilistic networks from data[J]. Machine learning, 1992(9): 309-347
[18] WU J, ZHOU R, XU S, et al. Probabilistic analysis of natural gas pipeline network accident based on Bayesian network[J]. Journal of Loss Prevention in the Process Industries, 2017, 46: 126-136.
[19] 于涛. 基于故障树贝叶斯网络的TDS-8SA顶驱装置故障分析与诊断[D]. 大连: 大连海洋大学, 2022.
[20] DE V J, KRAAK M H S, SKEFFINGTON R A, et al. A Bayesian network to simulate macroinvertebrate responses to multiple stressors in lowland streams[J]. Water Research, 2021, 194: 116952.
