00732424

Component

https://doi.org/10.3141/1556-15

http://scholar.google.com...neja&publication_year=1996

Accurate and reliable estimates of trip production of a study area are important for an accurate forecast from the four-step travel demand forecasting procedure. In the trip generation step, trip production estimates are considered more accurate, and trip attractions are adjusted while keeping the productions constant. This means that more accurate trip production rates will result in more reliable forecasts. Improving the accuracy of forecasts requires an extensive and reliable data base or improvement in the modeling techniques. Since data base enhancement is costly and time-consuming, an alternative methodology is proposed and examined for trip production prediction using artificial neural network (ANN) concepts and techniques. The data base used was made available by the Delaware Department of Transportation. The data were collected for 60 sites throughout Delaware between 1970 and 1974 and are based on field counts and home interviews. Twenty-six regression models were calibrated on these data. In addition, 18 ANN architectures were developed, and their predictions were compared with those from regression models. Comparisons indicate that the ANNs have the capability to represent the relationship between the trip production rate and the independent variables more accurately than regression analysis at no additional cost of increasing the data base.

This paper appears in Transportation Research Record No. 1556, Transportation Forecasting: Short-Term Practical Improvements, Travel Behavior Models and Issues, and Artificial Intelligence.

English

p. 131-136

1996

Transportation Research Record

0309059550

Figures (2) ; References (17) ; Tables (1)

Accuracy; Alternatives analysis; Estimates; Forecasting; Neural networks; Regression analysis; Travel demand; Trip generation

Data and Information Technology; Highways; Planning and Forecasting; I72: Traffic and Transport Planning

TRIS, TRB

Feb 19 1997 12:00AM

• ANALYSIS OF STATED PREFERENCES FOR INTERMODAL BICYCLE-TRANSIT INTERFACES
• ASSIGNING TRAFFIC OVER ESSENTIALLY-LEAST-COST PATHS
• COMPARABILITY AND TRANSFERABILITY OF COMMUTER BEHAVIOR CHARACTERISTICS BETWEEN CITIES: DEPARTURE TIME AND ROUTE-SWITCHING DECISIONS
• CONGESTION-BASED PEAK SPREADING MODEL
• FORECASTING BALANCED HIGHWAY VOLUMES USING MODELING PACKAGES AND SPREADSHEETS
• FUZZY FEEDBACK CONTROL FOR REAL-TIME DYNAMIC TRAFFIC ROUTING: USER EQUILIBRIUM MODEL FORMULATIONS AND CONTROLLER DESIGN
• HOUSEHOLD ACTIVITIES, LIFE CYCLE, AND ROLE ALLOCATION
• IMPLEMENTATION OF "HIGHWAY CAPACITY MANUAL"-BASED VOLUME-DELAY FUNCTIONS IN REGIONAL TRAFFIC ASSIGNMENT PROCESS
• IMPROVED MODELING OF NON-HOME-BASED TRIPS
• INCORPORATING WORK TRIP ACCESSIBILITY IN NONWORK TRIP GENERATION MODELS IN SAN FRANCISCO BAY AREA
• MULTIVARIATE MARGINAL FREQUENCY ANALYSIS OF ACTIVITY AND TRAVEL PATTERNS IN FIRST FOUR WAVES OF PUGET SOUND TRANSPORTATION PANEL
• RESIDENTIAL LOCATION, EMPLOYMENT LOCATION, AND COMMUTER RESPONSES TO PARKING CHARGES
• TRAVELER RESPONSE TO DAMAGED FREEWAYS AND TRANSPORTATION SYSTEM CHANGES FOLLOWING NORTHRIDGE EARTHQUAKE
• TRIP END MODE SPLIT
• TWO-STAGE VEHICLE AVAILABILITY MODEL