Questions Regarding Scoring Parameters and Calibration Process

[ArezHK]

Hi,
I have some questions regarding scoring parameters and the calibration process and would appreciate your assistance.

In the scoring module, there are different parameters, and as I learned about them and also examined them in various available MATSim scenarios, I encountered some doubts, which I explain below:

A. marginalUtilityOfDistance_util_m
This parameter is set to “0” in most of the existing scenarios, but in your courses, you mentioned it can be used for bicycles and should be set to zero for other modes. How is this parameter used, and what values are allowed?

B. marginalUtilityOfTraveling_util_hr
The default value for this parameter is “-6.” However, in the existing scenarios, it has different values: for car, it is set to “0,” and for bike, public transportation, and walking, various values have been used (e.g., for bike: 0, -5, -12.7, and -3; for public transportation: 0, -10.5, -3, and -2; and for walking: 0, -1 and -1.5). How are these values calculated, or why are they set as such?

C. marginalUtilityOfMoney
I assume this is related to the currency in Europe, which should be approximately “1.” However, in the Berlin scenario (5.5, 10%), it is set to 0.6. What is the reason for this?

D. performing
By default, this parameter is set to +6, but in one case, it is set to 6.88. Is it advisable to change this value, or should it remain at 6? How can we logically determine this number?

Additionally, I have some general questions about the calibration process:

1. Calibrating based on real life Modal Share Values
   If I want to calibrate my model based on modal share values, after setting the correct values for the above parameters, should I simply manipulate and test different values for the ASC of the available modes?
   o Assuming I have car, public transport (PT), and bike as routed modes, should I also consider the “ride” mode or just remove it from modes?
   o Are there any suggestions or relationships between different ASC values for different modes to reach the desired modal share more efficiently? (For example, I notice that in your existing scenarios for ASC_car, values are around -0.5 to -1.5, and for bike, they are around -1.5.)

2. Calibration Without Innovations
   Should the calibration process and the determination of ASC values be done without time, route, and mode innovations?

3. Scaling the Scenario
   If we have a population file consisting of only 10% of the real population of the area (e.g., 10,000 plans for a city with 100,000 residents), is it correct to set the flowCapacityFactor and storageCapacityFactor to 0.1? If so, what else should be considered? For instance, how should the capacity of public transportation (e.g., 50 seats per vehicle) be handled in the scaled-down scenario?

Thank you in advance for your assistance!