Lecture_23.mode_and_route_choice.extende

Lecture_23.mode_and_route_choice.extende - The Pennsylvania...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: The Pennsylvania State University Department of Civil and Environmental Engineering Lecture 24 – Trip Generation, Mode Choice and Traffic Assignment CE 321: Highway Engineering Spring 2008 Travel Demand Introduction Traffic forecasts for new construction Pavement Design Geometric Design (number of lanes, shoulder widths, etc.) Traffic forecasts for operational improvements Estimate effectiveness of alternative improvement projects Considerations Overall regional traffic growth/decline Provide sufficient level of service (LOS) Economic factors Traffic-generating activities (work & shopping) Spatial distributions Traffic diversion Network or system effects Traveler Decisions Temporal decisions When to travel (e.g. what time) How often to travel (e.g. frequency per week) Destination decisions Modal decisions (car, bus, walking, etc.) Spatial or route decisions (origin & destination) Traffic Forecasting Overview of Traffic Estimation Transportation Planning Process Land Use Trip Generation* Trip Distribution Modal Split* Traffic Assignment* Land Use Descriptors Character Intensity Location Trip Generation The purpose of trip generation estimation is to determine the number of trips to and from activities in an analysis area based on several different factors. Household trips instead of individual trips are modeled. Trip Generation Types Models seek to predict number of trips per hour or per day. Work Trips Shopping Trips Social/recreational Trips Typical Trip Generation Model Ti = b0 + b1 z1i + b2 z 2i + ... + bk z ki Ti = number of vehicle-based trips of given type in some specified time-period made by household i; zki = characteristic k of household i; bk = coefficient estimated from data for characteristic k. Factors Affecting Trip Generation Household Income Household Size and Composition Automobile Ownership Availability of Transit Density of Development Trip Generation Example 1 Number of peak-hour vehicle-based shopping trips per household = 0.12 + 0.09 (HH size) + 0.011 (HH income (1000’s)) – 0.15 (employment in HH neighborhood (100’s) HH has 6 members and annual income of $50,000. They currently live in neighborhood with 450 retail employees, but are moving to neighborhood with only 150 retail employees. Calculate number of vehicle-based trips made before and after move. Trip Generation Solution 1 Before Move: # trips = 0.12 + 0.09(6) + 0.011(50) – 0.15(4.5) # trips = 0.535 peak-hour trips After Move: # trips = 0.12 + 0.09(6) + 0.011(50) – 0.15(1.5) # trips = 0.985 peak-hour trips Trip Generation Example 2 A neighborhood has 205 retail employees and 700 households with following characteristics: HH size 2 3 3 4 Income $40,000 $50,000 $55,000 $40,000 Non-workers Workers in peak-hour departing 1 2 1 3 1 1 2 1 Type 1 2 3 4 Trip Generation Example 2 There are 150 type 1, 200 type 2, 300 type 3, and 100 type 4 households. # shopping trips = 0.12 + 0.09 (HH size) + 0.011 (HH income (1000’s)) – 0.15 (employment in HH neighborhood (100’s) # social/recreational trips = 0.04 + 0.018(HH size) + 0.009 (HH income (1000’s)) + 0.16 (# non-working HH members) Trip Generation Example 2 Find: Total peak hour trips Method: Find # trips by purpose for each HH type Multiply by # HH of each type Add trips by type Type 1: 0.12+.0.09(2)+0.011(40)-0.15(2.05)= 0.4325 trips/HH Shopping trips For 100 HH this yields 43.25 trips in peak period for shopping Type 2: 0.12+.0.09(3)+0.011(50)-0.15(2.05)= 0.6325 trips/HH For 200 HH this yields126.5 trips Trip Generation Example 2 Total shopping trips: 441 vehicle based shopping trips in peak hour Using similar approach for social recreational trips: 542 vehicle-base social recreational trips in peak hour Work Trips: (Workers departing) x (#HH) for each type 1(100)+1(200)+2(350)+1(50) = 1050 vehicle-based work trips in peak hour Total peak hour trips: 441+542+1050=2033 This is a simplified approach, but gives you an idea of some of the influential factors we Mode Choice Mode Choice/Split Analysis allows the planner to determine the magnitude of travel by individual modes. Mode Choice Models The choice of mode is a decision that is akin to the probability of doing something or not doing something. A model that will estimate probabilities based on measurable variables is needed. Logit Model (maximizes utility value) Utility, a microeconomic term, can be defined as the capacity to satisfy wants. Logit Model Uim = Σ bmk zimk i Uim = specifiable portion of utility of alternative m for traveler i; bmk = coefficient estimated from data for mode/destination alternative m corresponding to mode/destination or traveler characteristic k; zimk = traveler of mode/destination characteristic k for mode/destination alternative m for traveler i. Logit Model Pim = ∑e s e U im U sm Pim = probability that traveler i selects alternative m. Uim = specifiable portion of utility of alternative m for traveler i; Mode Choice Example Mode choice utility model from small urban area is shown below: UDL = 2.2 – 0.2(costDL) – 0.03(TTDL) USR = 0.8 – 0.2(costSR) – 0.03(TTDL) UB = -0.2(costB) – 0.01(TTB) Where DL = automobile-drive-alone; SR = automobile-shared-ride; B = bus. Between a residential area and an industrial complex, 4000 workers depart for work during peak hour. Cost of driving automobile is $4.00 with a TT of 20 minutes – the bus fare is $0.50 with a TT of 25 minutes. If the shared-ride options always consists of two travelers sharing costs equally, how many workers will take each mode? Modal Choice Example UDL = 2.2 - 0.2 (4) - 0.03 (20) = 0.8 USR = 0.8 - 0.2 (2) - 0.03 (20) = -0.2 UB = -0.2 (0.5) – 0.01 (25) = -0.35 Modal Choice Example PDL = e0.8 / [e0.8 + e-0.2 + e-0.35] = .594 PSR = .819 / 3.749 = 0.218 PB = 0.705 / 3.749 = 0.188 Multiplying probabilities by 4000 gives: 2380 drive alone 870 share ride 750 bus Because of gas prices increases, cost of drive alone travel increases by $2.00 Is there an increase cost for sharing a ride? Extended Example Reflecting Gas Price Increase Because of gas prices increases, cost of drive alone travel increases by $2.00 Is there an increase cost for sharing a ride? Extended Example Reflecting Gas Price Increase YES; assume half cost of DR as before. Is there a cost increase for transit? Because of gas prices increases, cost of drive alone travel increases by $2.00 Is there an increase cost for sharing a ride? Extended Example Reflecting Gas Price Increase YES; assume half cost of DR as before. Is there a cost increase for transit? YES; diesel fuel increase Assume fare increases from 50 cent to 55 cents New Mode Shares UDL = 2.2 - 0.2 (6) - 0.03 (20) = 0.4 USR = 0.8 - 0.2 (3) - 0.03 (20) = -0.4 UB = -0.2 (0.55) – 0.01 (25) = -0.36 Mode shares: PDL = 0.52 (old = 0.59) PSR = 0.23 (old = 0.22) Importance of Logit Model Estimates mode shares at disaggregate level – the individual traveler Model includes affect of variables we might want to change to influence mode share Travel time Cost of transit Cost of auto travel Other variables typically included: Demographics of travelers Frequency or waiting time of transit service Route Choice Result is traffic flow (vehicles per hour) on specific highway routes. Route choice is function of travel times and traffic flow. Need relationship between route travel time and route traffic flow (highway performance function). Highway Performance Function Route travel time Free-flow travel time 0 Traffic flow User Equilibrium Assume travelers select route between origins and destinations based on route travel times only. Assume travelers know travel times that would be encountered on all possible routes between origin and destination. Travelers will select the route that minimizes personal travel time. Route Choice Example Two routes connect a city and a suburb. See speed limit and route length data below. Studies show that total TT on route 1 increases two minutes for every additional 500 vehicles added. Minutes of TT on route 2 increase with the square of the number of vehicles (1000’s). Determine user equilibrium travel times. Route 1 2 Speed Limit (mph) 60 45 Length (miles) 6 3 Route Choice Example Solution Free-flow travel times: Route 1 – 6 miles / 60 mph = 0.1 hr. = 6 min. Route 2 – 3 miles / 45 mph = 0.0667 hr. = 4 min. t1 = 6 + 4x1 t2 = 4 + 4x22 t1, t2 = average travel time on routes 1 & 2 (min) x1, x2 = traffic flow on routes 1 & 2 (1000’s vph) Performance Functions Route Choice Example Solution Note: At flows above q’, route 2 is congested enough that route 1 is a viable option. ...
View Full Document

Ask a homework question - tutors are online