基于决策树支持向量机算法的船用设备故障诊断

目的 提高船用设备的智能化水平,增强船舶的安全性、可靠性,对船上设备进行状态监测,并基于监测数据对设备健康状况进行评估,对可能存在的故障工况进行识别.方法 通过采集机舱内的振动数据,对数据进行预处理、快速傅里叶变换,提取1/3倍频带特征,将倍频带谱信号作为特征向量,利用支持向量机算法进行模型训练及分类.对于船上多种工况及可能存在的多种故障类别,采用决策二叉树方法,提出一种快速、准确的状态监测及故障诊断策略.结果 在实验室工况下识别准确率接近100％.结论 该方法能够对船用设备进行状态监控、故障诊断、健康评估等提供支持,为设备检修、传感器布置等决策提供依据.

Research on Marine Equipment Fault Diagnosis Based on Decision Tree Support Vector Machine Algorithm

In order to improve the intelligent level of marine equipment and enhance the safety and reliability of ships, it is urgent to monitor the status of the equipment on the ship, and evaluate the health status of the equipment based on the monitoring data to identify possible fault conditions. The vibration data collected in the cabin was preprocessed using fast Fourier transform to extract the one-third octave band features, and the octave band spectrum signal was used as the feature vector. Support vector machine algorithm was used for model training and classification. For a variety of working conditions on the ship and possible multiple fault categories, the decision binary tree method was used to propose a fast and accurate state monitoring and fault diagnosis strategy. The identification accuracy under laboratory conditions was close to 100%. This method can provide support for state monitoring, fault diagnosis and health assessment of marine equipment, and provide basis for decision-making of equipment maintenance and sensor arrangement.