Trip Interchange Model ❑ The trip interchange model considers system level of service variables: ▪ In-vehicle travel time ▪ Excess (i.e., walking and waiting) time ▪ Travel costs ▪ Economic status of trip makers Trip Interchange Model Note the inverse relationship; I transit is numerator to determine proportion of auto travel Trip Interchange Model Example ❑ Suppose the data shown in table 12.21 show travel data between a suburban zone S and a downtown zone D. The exponent value for the impedance factors is 2.0 and the median income is $24,000. Determine the expected percentage of auto and transit trips. Trip Interchange Model Example Trip Interchange Model Example
▪ Therefore, 58.3 percent of travel between the zones S and D is
expected to be via transit. Quiz!
Consider Air transportation between Boston and New
York. What are the level-of-service variables that you would consider to be important? What measures you use in evaluating this service? ▪ Travel time ▪ Type of Aircraft ▪ Frequency of flights (average waiting time..?!) ▪ Access time ▪ Fare ▪ Comfort Level of service variable
Level –of-service variables and units
Variables Units Travel time tt Minutes Access time ta Minutes Waiting time tw Minutes Fare F $ Comfort H “Hugs” Level of service variable
You have several variables and units. We need to measure the
level of service in unidirectional way. We need a way of collapsing these variables into a single variable.
❑ So, we define a measure called “utils” for “utility”. We
can convert all those level of service variables measured in different units, to utils. Level of service variable
Let’s define a variable U, which is the utility of a travelers’
choice in utils: (in some references it is called V instead)
U = a0 + a1 tt + a2 ta + a3 tw + a4 F + a5 H
❑ Coefficient a1, …, a5 help to convert all the units to utils. a0 is in
utils, and it is a constant that use to calibrate the utility equation.
❑ Coefficients can be negative or positive.
Mode Choice & Utility
Suppose we have three possible modes of travel from Boston to
New York – air, train, auto- and you could measure each of the level-of-service variables for each mode. We could compute the utility of each mode,
Uair , Utrain , Uauto
The mode with highest utility is one you would choose! Mode Choice & Utility
Much modern in utility theory uses a probabilistic approach. For
example, the probability a traveler selects the air mode is as follows:
𝑈𝑎𝑖𝑟 Pair = 𝑈𝑎𝑖𝑟+𝑈𝑡𝑟𝑎𝑖𝑛+𝑈𝑎𝑢𝑡𝑜
𝑒 𝑈𝑎𝑖𝑟 Pair = 𝑒 𝑈𝑎𝑖𝑟 + 𝑒 𝑈𝑡𝑟𝑎𝑖𝑛 + 𝑒 𝑈𝑎𝑢𝑡𝑜 • The probabilistic approach is intended to reflect the fact that people have different utilities, and perhaps we have not captured all LOS variables in our formulation. Mode Choice & Utility
❑ Mode Choice Models are used to try to predict travelers mode choice ❑ Contemporary models are based on using UTILITY or DISUTILITY functions
❑ These functions are meant to express the level of satisfaction (for
utility functions) or dissatisfaction (for disutility functions) with a given mode
❑ Once the utility function is calculated for each mode, the probability that a given mode will be chosen can then be calculated Mode Choice & Utility
Some of factors the can affect mode choice include:
1. Type and purpose of trip
2. Car ownership status 3. Cost (mostly out of pocket cost) 4. Door-to-door travel time 5. Convenience/service/comfort 6. Prestige 7. Availability 8. Accessibility of mode 9. Land use characteristics of start and end point
❑ Obviously, not all of these factors can be effectively incorporated into
a quantitative model of mode choice. We need to be cognizant of those factors that are important in influencing choice but are not fully accounted for in mode choice models. Utility function
In general a utility function takes the following form:
uk = ak + a1 X1 + a2 X2 + ….. ar Xr + ε0
Where
uk – utility function for mode k
ak – modal constant Xr – variables measuring modal attributes such as cost or time of travel ar – coefficient associated with each attribute ε0 – error term Logit Models ❑ The logit model considers the relative utility of each mode based on various modal attributes.
where Logit Models ❑ The choice of a mode is expressed as a probability distribution. ❑ If auto (A) and Transit (T) modes are being considered, the probability of selecting the auto mode A can be written as: Mode Choice & Utility
If utility function, uk, is assumed to be a Weibull Probability
Distribution then the Multinomial Logit Model is used to calculate the probability that a traveler will choose a given mode.
Multinomial Logit Model:
p(k) = e uk /Σ e ux
Hint: The Weibull distribution is a continuous probability distribution named after Swedish mathematician Waloddi Weibull. He originally proposed the distribution as a model for material breaking strength, but recognized the potential of the distribution in his 1951 paper A Statistical Distribution Function of Wide Applicability. Today, it’s commonly used to assess product reliability, analyze life data and model failure times. The Weibull can also fit a wide range of data from many other fields, including: biology, economics, engineering sciences, and hydrology (Rinne, 2008).
Mode Choice & Utility The number e is an important mathematical constant that is the base of the natural logarithm. It is approximately equal to 2.71828, and is the limit of (1 + 1/n)n as n approaches infinity. It can also be calculated as the sum of the infinite series.
The natural log at (x-axis) e, ln(e), is equal to 1
e = 2.7182818284 5904523536 0287471352 6624977572 4709369995
Example: The mode available between Zone I and J are:
(i) Passenger car (PC), (ii) Bus Find the market share for each mode given the attribute table for the modes.
o Suppose the utility function is as following:
uk = ak – 0.025 X1 – 0.032 X2 – 0.015 X3– 0.002 X4
Where x1 – access plus egress time (min) x2 – waiting time (min) x3 – line haul time (min) x4 – out of pocket cost (cents) ε0 – error term Mode Choice & Utility
The attribute table for each mode is given below:
x1 x2 x3 x4 PC 5 0 20 100 aPC = +0.00 Bus 10 15 40 50 aBus = -0.10
Solution: U(A) = -0.625 U(B) = -1.530
Probability of selecting PC,
p(A) = e(-0.625) / [e(-0.625)+e(-1.530)] = 0.71
Quiz!
The mode available between City A and City B are: Passenger car (PC), Bus, Train, and Airplane. Find the market share for each mode given the attribute table for the modes. Suppose the utility function is as following: Uk = ak – 0.015 X1 – 0.025 X2 + 0.03 X3 – 0.002 X4 + ε0 Where X1 – Travel time (min) X2 – waiting time (min) LOS Variables X3 – Comfort (Hugs) X4 – out of pocket cost (cents) Available Mode X1 X2 X3 X4 ε0 – error term (Negligible) Passenger car 12 0 40 100
aPC = 0.00 Bus 15 25 20 50
aBus = -0.10 Train 10 15 25 40 aTrain = -0.10 Airplane 5 30 30 150 aAirplane = -0.01 Logit Model Example ❑ Data has been obtained for travel between an additional suburban and downtown zone. The logit model has been developed to estimate the mode share between private automobile and public transit between these two zones:
where Logit Model Example ❑ Travel and cost data for each mode is provided in the following table. Logit Model Example ❑ Use the logit model to determine the percent of travel in the zone by auto and transit.
[Ebooks PDF] download Introduction to Integral Calculus Systematic Studies with Engineering Applications for Beginners 1st Edition Ulrich L. Rohde full chapters
[Ebooks PDF] download Introduction to Integral Calculus Systematic Studies with Engineering Applications for Beginners 1st Edition Ulrich L. Rohde full chapters
