In this demonstration, you can perform a pursuit tracking task, analyze your performance and find a control model that fits your performance best.
The pursuit tracking task consists of using your mouse to follow the red target bars with the green cursor bar. The task lasts for 60 seconds (the first 5 seconds are not recorded). The controlled variable is the distance between the cursor and the target, and you need to keep it as close to zero as you can.
Select the degree of difficulty: 1 2 3 4 5
Analysis
The graph below shows your tracking performance, and bellow it you can see the RMS error between the target and the cursor.
Tracking results
Tracking (distance) RMS:
Model fitting
The first graph shows the difference between modeled cursor and real cursor. You can change the parameters of the model by clicking and dragging them left to decrease and right to increase. Parameters of the model that best describe your tracking performance are those that have smallest RMS difference. You can also click the Auto Fit button to get the parameters calculated automaticaly. The 'Ko' parameter is loop gain, 'delay' is the input transfer lag in milliseconds, 'ref' is the reference distance between cursor and target, and 'damping' is related to the output slowing factor. The second graph shows model tracking performance in tracking the target.
Cursor vs model cursor RMS difference:
Model tracking (distance) RMS: