基于SVM模型算法和大数据分析技术的船舶设备故障诊断

Big Data Analysis with SVM Model Algorithm for Fault Diagnosis of Marine Equipment

为在船舶设备发生故障时能准确、及时地定位故障发生根源,保证船舶安全、经济运行,采用大数据分析方法和支持向量机(Support Vector Machine,SVM)模型算法对船舶设备进行故障诊断,提前预测可能发生的故障。以船舶柴油机滑油压力低故障为例,应用Python语言,通过SVM模型算法预测该故障的发生概率。结果表明,在已采集的船舶数据样本的训练集和测试集上,数据拟合和故障预测的效果十分理想,预测故障发生的准确率较高。

The SVM(Support Vector Machine)model algorithm for predicting the probability of low lubrication oil pressure in a diesel engine is developed with Python.The model is trained with a training data set and verified with a test data set.Satisfactory data fitting and fault prediction are demonstrated.