为适应无人艇高可靠性和智能运维保障的发展需求，本文提出一种基于改进VIT神经网络模型的无人艇电力系统故障诊断方法，通过改进的注意力机制对模型参数进行求取，降低算法空间的复杂度，且避免模型参数限于局部最优。通过对交流低压无人艇电力系统进行短路故障仿真建立故障数据集，采用连续小波变换对故障电压序列数据进行特征提取，该特征数据用于训练改进VIT模型，实现无人艇电力系统故障诊断。对改进VIT模型与卷积神经网络CNN、深度收缩残差网络DRSN、VIT模型的故障识别性能进行仿真对比研究，结果表明改进VIT模型具有较高的故障诊断准确度，且受短路故障电阻的变化影响小、适应性更强。
In order to meet the development demands for high reliability and intelligent maintenance support of unmanned boats, a fault diagnosis method of unmanned boat power systems based on an improved VIT neural network model is proposed. By employing an enhanced attention mechanism to determine model parameters, the algorithmic space complexity is reduced, and the model parameters are prevented from being confined to local optima. Through short-circuit fault simulations of AC low-voltage unmanned boat power systems, a fault dataset is established. Continuous wavelet transform is utilized to extract features from fault voltage sequence data. These feature data are used to train the improved VIT model to achieve fault diagnosis for unmanned boat power systems. The paper conducts a comparative simulation study on the fault recognition performance of the improved VIT model,convolutional neural networks (CNN), deep residual shrinking networks (DRSN), and the standard VIT model. The results indicate that the improved VIT model exhibits higher fault diagnosis accuracy and is less affected of changes in short-circuit fault resistance, demonstrating stronger adaptability.
2024,46(18): 154-158 收稿日期：2024-3-4
DOI：10.3404/j.issn.1672-7649.2024.18.027
分类号：TP206+.3；U672.7+1
基金项目：江苏省科技成果转化专项资金项目（BA2022066）
作者简介：郑海山(2000-),男,硕士研究生,研究方向为船舶智能运维
参考文献：
[1] 贾哲宇, 温华兵, 朱军超, 等. 基于随机森林方法的柴油机涡轮增压器故障诊断[J]. 舰船科学技术, 2023, 45(6): 109-13.
JIA Zheyu, WEN Huabing, ZHU Junchao, et al. Fault diagnosis of diesel engine turbochargers based on random forest method[J]. Ship Science and Technology, 2023, 45(6): 109-13.
[2] 杨奕飞, 冯静. 基于HMM-SVR的船舶动力设备故障模式识别与状态预测研究[J]. 船舶工程, 2018, 40(3): 68-72+97.
YANG Yifei, FENG Jing. Fault pattern recognition and state prediction research of ship power equipment based on HMM-SVR[J]. Ship Engineering, 2018, 40(3): 68-72+97.
[3] YIFEI Y, MINJIA T, YUEWEI D, et al. An improved CS-LSSVM algorithm-based fault pattern recognition of ship power equipments[J]. Plos One, 2017, 12(2): e0171246.
[4] 高际航, 张艳. 基于FWA-PSO-MSVM的船舶区域配电电力系统故障诊断[J]. 计算机科学, 2022, 49(S2): 956-60.
GAO Jihang, ZHANG Yan. Fault diagnosis of shipboard zonal distribution power system based on FWA-PSO-MSVM[J]. Computer Science, 2022, 49(S2): 956-60
[5] 陈世鹏, 杨奕飞, 张林, 等. 基于CEEMDAN和SCA-MRVM的滚动轴承故障诊断[J]. 计算机科学, 2021(10): 53-9.
[6] SENEMMAR S, ZHANG J. Deep learning-based fault detection, classification, and locating in shipboard power systems[Z]. 2021 IEEE Electric Ship Technologies Symposium (ESTS). 2021: 1-6.10. 1109/ests49166.2021. 9512342
[7] 徐舒玮, 邱才明, 张东霞, 等. 基于深度学习的输电线路故障类型辨识[J]. 中国电机工程学报, 2019, 39(1): 65-74+321.
XU Shuwei, QIU Caiming, ZHANG Dongxia, et al. A Deep learning approach for fault type identification of transmission line[J] Proceedings of the CSEE, 2019, 39 (1): 65-74+321
[8] 杨奕飞. 舰船装备健康评估与管理若干关键技术研究[D]. 南京: 南京理工大学, 2019.
[9] DOSOVITSKIY A, BEYER L, KOLESNIKOV A, et al. An image is worth 16x16 words: transformers for image recognition at scale[J/OL] 2020, arXiv: 2010.11929[https://ui.adsabs.harvard.edu/abs/2020arXiv201011929D.10.48550/arXiv.2010.11929
[10] VASWANI A, SHAZEER N, PARMAR N, et al. Attention Is All You Need[J/OL] 2017, arXiv: 1706.03762[https://ui.adsabs.harvard.edu/abs/2017arXiv170603762V.10.48550/arXiv.1706.03762
[11] LI Q, LUO H, CHENG H, et al. Incipient fault detection in power distribution system: a time–frequency embedded deep-learning-based approach[J]. IEEE Transactions on Instrumentation and Measurement, 2023, 72: 1-14.
