CS452 - Real-Time Programming - Spring 2008

Lecture 29 - Practical Control and Calibration


Questions & Comments

  1. How to give a demo

Practical Control

Data

What data do you have?

How do you find out where the train is?

We need to use the data to get a good estimate of the velocity, which may be a function of

This data has three types of errors

  1. screw-up errors
    1. throw them out
    2. sometimes you can eliminate them
  2. random errors
  3. systematic errors

How useful is yesterday's data?

Eliminating screw-up errors

Redefine the track

For example, if a sensor malfunctions frequently

Transforming random errors

You can sometimes identify patterns in what you think are random errors

Subdivide the data.

Processing Data

Averaging

Subdivide data into cells, average within cells.

Averaging techniques

  1. whole series: average' = average * (n-1)/n + data / n
  2. moving window, varying

    equal weighting: average' = average + (new-data - dropped-data)/n

Extrapolating


Return to: