A method for automatically shifting a base line, comprising:
- collecting processing data in time series and predictive maintenance (PM) data in time series by a data collection unit;
- the collected processing data and the predictive maintenance (PM) data being inputted to a base line process unit through a data interface unit, and inserting the predictive maintenance (PM) data into the processing data for figuring out a processing section before predictive maintenance by a predictive maintenance section, and a processing section after predictive maintenance by the base line process unit;
- calculating an original mean value and an original standard deviation for each section of the processing section before predictive maintenance by the predictive maintenance section, and the processing section after predictive maintenance by the base line process unit;
- filtering outlier data, and calculating a first mean value and a first standard deviation for each section of the processing section before predictive maintenance by the predictive maintenance section, and the processing section after predictive maintenance by the base line process unit;
- calculating a difference between value of each data point of each section of the processing section before predictive maintenance by the predictive maintenance section, and the processing section after predictive maintenance and the first mean value of each section of the processing section before predictive maintenance by the predictive maintenance section, and the processing section after predictive maintenance, and ranking the data points of each section of the processing section before predictive maintenance by the predictive maintenance section, and the processing section after predictive maintenance depending on absolute value of the difference by the base line process unit;
- selecting front N % data points of each section of the processing section before predictive maintenance by the predictive maintenance section, and the processing section after predictive maintenance, and calculating a second mean value and a second standard deviation for the front N % data points of each section of the processing section before predictive maintenance by the predictive maintenance section, and the processing section after predictive maintenance by the base line process unit;
- filtering outlier data, and calculating a third mean value and a third standard deviation for each section of the processing section before predictive maintenance by the predictive maintenance section, and the processing section after predictive maintenance, wherein the mean value of each section is a base line of each section of the processing section before predictive maintenance by the predictive maintenance section, and the processing section after predictive maintenance by the base line process unit;
- and
- shifting and aligning the base lines of the processing section before predictive maintenance by the predictive maintenance section, and the processing section after predictive maintenance by the base line process unit.