Virtual V8 Track Mapping

For motor racing applications, track mapping involves using the vehicle data logger and associated instrumentation as a simple inertial navigation system. The principle is best described as having a constantly varying velocity vector, with data for this recorded for every time increment of the data logger.

The data logger velocity channel gives the magnitude of this velocity vector while the data logger lateral acceleration channel gives the basis for the direction of the velocity vector.

Errors in the instrumentation readings can propagate rapidly, so very high quality instrumentation is required for best results. Typical instrumentation will give good results though there will be an appreciable error present. The track map generated will generally define a consistent set of data, but will probably not result in a dimensionally correct track layout.

This image shows the raw data logger information for a lap of Phillip Island. This raw data analysis gives a track length of 4312 metres. The actual lap length is 4450 metres. This indicates a scaling error is present within the velocity data recorded. The track path does not close due to instrumentation errors.

This image is the logger information corrected for lap distance. This has been corrected to obtain the revised map shown. Note how applying this scaling factor has changed the entire shape of the map.

This image is a map from the raw logger data, but with an offset added to the lateral acceleration values. This has not yet had the distance correction applied.

The lateral accelerometer provides the heading value. This accelerometer can suffer from errors due to any minor misalignment in the car and also from scaling errors. Adding an offset value will compensate for the misalignment error. This value can usually be determined by trial and error.

This image is a trace with both the distance and offset corrections applied. The lateral acceleration offset here has a value of 0.0105 and is applied first, then the distance correction is applied.

Changing the acceleration values after the distance correction could result in a change in the calculated track length, so the distance correction is applied last, in order that the true track length is obtained for the final data set.

This image is a trace with a smaller lateral acceleration offset value of 0.00315 applied. As can be seen when comparing this map with the preceeding one, this map is much closer to the result where the track path closes. Having the lap start and end at the same point is the objective here. The lap should also ideally have the same heading for the start and end locations in order to obtain a consistent data set. Iterating to a final value can be employed to obtain a close approximation to the actual accelerometer offset value.

Note that no allowance has been made here for vertical variation in the track profile, so a perfect result should not be expected for that reason alone. Some combination of velocity correction (scaling factor) and lateral accelerometer correction (offset and scaling factor) should lead to a nearly closed track map.