Use Telco OD Matrices for secondary location choice

[tduberne]

We have OD matrices between municipalities in Switzerland, from Monday to Friday, per month of the year, per hour of the day, for one continuous year of collected data. That would complement the counts nicely to allocate secondary locations.

We have it as an absolute number of observation (not scaled to the whole population), but assuming a uniform market share across the country, relative flows should be usable. Fine for us, as we do not generate or delete trips.

• [ ] Extend location choice to include OD flows as target
• [ ] Generate scenario using yearly averages

Possible extensions: use the classification per month to generate one winter and one summer scenario.

[WolfgangScherr]

Hello Tibault, Very interesting. I would be interested to check the quality of the data: is the coverage homogenous or biased; has the imputation of modes succeeded, do the trip length distributions fit other empirical ones, are the network loads plausible, ... Do you plan to do quality checks? Wolfgang