摘要
由于船用发电机组结构复杂,运行产生的数据较多且复杂性较高,给发电机的运行异常识别带来严峻挑战,因此研究基于多特征融合的船用发电机运行异常识别方法。采集并预处理船用发电机的运行数据,从运行数据中提取不同类型与维度的船用发电机运行状态特征,对提取到的多特征进行融合处理,采用SVM分类识别融合后的特征,得到船用发电机运行异常识别结果。实验结果表明,设计方法识别船用发电机运行异常的准确度为97.68%,验证了该方法的有效性与可行性。
Complex structure of marine generator sets and large and complex data generated during operation pose a serious challenge for identifying abnormal operation of generators.Therefore a method for identifying abnormal operation of marine generators based on multi feature fusion was studied.The operation data of marine generators were collected and preprocessed.Different types and dimensions of marine generator operation status features were extracted from the operation data and fused.By using SVM classification,the fused features were identified,obtaining the abnormal operation recognition results of marine generators.Experimental results show that the designed method achieves an accuracy of 97.68% in identifying abnormal operation of marine generators,and thus is effective and feasible.
作者
范大鸣
FAN Daming(Bohai Shipbuilding Vocational College,Huludao 125100,China)
出处
《电工技术》
2024年第5期51-53,共3页
Electric Engineering
关键词
多特征融合
船用发电机
运行异常
异常识别
multi feature fusion
marine generator
abnormal operation
abnormal identification
作者简介
范大鸣(1980-),硕士研究生,副教授,研究方向为电气自动化技术。
