This training highlights some techniques to preprocess your data in TrendMiner via 5 exercises.
- The first 2 exercises dive into data filtering via both static and dynamic filters. Applying filters lets you perform any further analysis on a specific part of your data, for example those periods where you are producing a specific product grade
- Exercises 3 to 5 show how to modify your data via the Tag Builder in several ways to, for example, compensate for disrupted sensor measurements or historian disconnections:
- Adding limits of a process value, removing outliers
- Filling in periods with most recent data instead of limiting the data
- Removing data and interpolating over the removed periods
Original training date: February 6 & 7, 2024
Recordings: