Modelling intercity bus passenger travel demand in Newfoundland
The purpose of this study was to develop a passenger forecasting model for intercity bus travel demand in Newfoundland. This involved the use of routinely published census data and mathematical techniques to predict and analyse passenger flow volumes. -- Modelling was performed at the trip generatio...
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
Memorial University of Newfoundland
1981
|
| Subjects: | |
| Online Access: | https://research.library.mun.ca/5283/ https://research.library.mun.ca/5283/1/Pilgrim_RobertWalter.pdf https://research.library.mun.ca/5283/2/Pilgrim_RobertWalter.pdf |
| Summary: | The purpose of this study was to develop a passenger forecasting model for intercity bus travel demand in Newfoundland. This involved the use of routinely published census data and mathematical techniques to predict and analyse passenger flow volumes. -- Modelling was performed at the trip generation and trip distribution stages of transportation planning. Trip generation was achieved using a library computer program for multiple linear regression analysis. During this stage, the most appropriate traffic zone size was assessed. Results were that an aggregation of census areas was proven more effective than single communities in relating passenger and demographic data. Trip distribution was carried out by applying the gravity model. A computer program, which included testing of the gravity model output, was developed and written by the author for this research. A statistical criterion called the chi-square-test was successfully applied in analyzing output of the gravity model. This test indicated the closeness between gravity model and observed passenger data and was judged superior to the visual method of determining when the gravity model is considered calibrated. The chi-square statistics were also applied in analysis of test year data. -- Testing of the model was achieved through comparison of predicted passenger data with data collected by actual counts during survey periods of four consecutive years following the base year. The model predicted a linear increase in ridership over the test years, whereas the actual data showed that fluctuations were present. Factors relating to dynamic change in the mode, which were included in the model, were accountable for deviations between predicted and observed data. |
|---|