may/june 2018 ieee power & energy magazine 63 figure 7. A framework for power equipment data cleaning. Temperature Air Pressure Altitude External Data Current/Voltage Active Power Reactive Power Power Grid Data Power Generation Condition Storage Condition Transformation Condition State Monitoring Data Nameplate Parameters Fault Records Maintenance Records Equipment Data Input Data Data Acquisition Cross-Platform Data Storage 0 0 100 200 300 400 500 600 100 Noise Data 50 150 Data Before Cleaning Data After Cleaning 200 300 Missing Data 250 Neural Networks/SVM Multiple Linear Regression Extreme Learning Machine Influencing Factors Mean Method Regression Method Multiple Imputation K-Nearest Distance Method Data Cleaning Abnormal Data Detection and Governance Missing Data Imputation Output Data