Fig 1.
The CPM with FFT DAQ System.
Fig 2.
The Basic architecture of the model.
Fig 3.
EDA architecture for CPM Dataset.
Table 1.
The 5-point summary of CPM dataset.
Fig 4.
Hypothesis testing of CPM dataset.
Fig 5.
Boxplot analysis of casing, impeller & bearing vibration.
Fig 6.
Pie chart of dependent variables.
Fig 7.
Bivariate analysis using line plot for casing, bearing & impeller vs pressure.
Fig 8.
Bivariate analysis using bar plot for pressure, current & voltage vs condition.
Fig 9.
Heatmap of CPM dataset.
Fig 10.
Impeller vs casing with bearing scatter.
Fig 11.
K-means clustering of bearing vs c_temp vs casing.
Fig 12.
Logistic Classifier Confusion Matrix.
Table 2.
Statistic Result of Logistic Regression Classifier.
Fig 13.
Naïve Bayes Classifier Confusion Matrix.
Table 3.
Statistic Result of Naïve Bayes Classifier.
Fig 14.
Confusion Matrix for Support Vector Classifier.
Table 4.
Statistic Result of Support Vector Classifier.
Fig 15.
RF Tree and Confusion Matrix for Random Forest.
Table 5.
Statistic Result of Random Forest Classifier.
Table 6.
Performance Table for ML Models.
Fig 16.
Performance & computational efficiency comparison of ML models.
Fig 17.
Training -validation accuracy & lost for ANN.