Adding Mobility Attitudes

[bsmoehring]

Add the mobility attitudes from previous research projects as personal attributes in MATSim model the influence of a possible shift in those attitudes. This will most likely be achieved by assigning specific scoring parameters to different subpopulations.

[wese-da]

I just had one additional idea concerning the scoring parameters:

We could experiment with the "cadyts approach" (maximum likelihood estimation) here:

• Score the modes with the default MATSim parameters
• Add an offset value (+ / -) that accounts for the mobility attitude (variable over the attitude groups)
• Mutate the offsets during simulation until the modal split of the mobility attitude groups fit the survey data

OR: alternatively to mutating the offsets we could try using the attitudes from Andreas' files and weight them with factor x, so that we have a score of the type:

U = V + x * attitude

[00Bock]

on hold. Results were promising but did not reach the expected quality levels for Osnabrück