Question

Assume population and HBW trip data for 5 TAZs are collected for the base year (Table 1) to calibrate a linear trip generation model as follows: HBW Trip = 100 + 1.9 Population.

To validate the model, we’re using 2017 data for only 3 TAZs (Table 2). Evaluate the prediction power of the model? How well it’s calibrated/validated?

Table 1

Table 2

TAZ	2020 Population	2020 HBW Trips
1	100	290
2	150	375
3	150	390
4	200	480
5	75	240

TAZ	2017 Population	2017 HBW Trips
1	80	230
2	120	310
3	140	330

Answer 1

It is a linear regression model of trip generation model of travel demand forecasting. In advice problem home based working trips is taken as a function of population of rhe traffic analysis zones (TAZ). To evaluate prediction power of the model let us assume the function as:

HBW Trips = 100 + A × Population

From table 1 Taking average of Population and HBW Trips and putting values in above equation we shall calculate the value of A:

(290+375+390+480+240)/5 = 100 + A × ((100+150+150+200+75)/5)

= 100 + A × 135 = 355

= 135A = 355-100

= 135A = 255

= A = 255/135

= A = 1.8888 1.9

So, putting value of A in above equation, the linear regression equation will be

HBW Trips = 100 + 1.9 Population

Hence validated