DEVELOPMENT AND TESTING OF A SET OF MATHEMATICAL 
MODELS FOR TRAVEL DEMAND ESTIMATION 
BY 
W.W.M.R.K WIJESUNDERA 
THIS THESIS WAS SUBMITTED TO 
THE DEPARTMENT OF CIVIL ENGINEERING 
OF THE UNIVERSITY OF MORATUWA IN PARTIAL FULFILLMENT OF 
THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ENGINEERING 
7-3 0 
©©•(CO© 8£s)ga. {§ 
»s\>op©. ' ^ 
DEPARTMENT OF CIVIL ENGINEERING 
UNIVERSITY OF MORATUWA 
SRI LANKA 
JUNE 2001 
073282 T C 
University of Moratuwa 
73282 
DECLARATION 
I certify that this dissertation does not incorporate without acknowledgement any material 
previously submitted for any other course in any university and to the best of knowledge 
and belief does not contain any material previously published or written or orally 
communicated by another person except where due reference is made in the text. 
Date: * a / W a © o / ^J^X™.?*?** 
Candidate's Signature 
(Mr.W.W.M.R.K.Wijesundera) 
Date: 
Project Supervisor's Signature 
(Dr. Amal S. Kumarage) 
ABSTRACT 
Regional traffic models are a useful tool in planning transport infrastructure in 
keeping with anticipated human settlement patterns and activities. The amount and nature 
of travel depend on the population size, income level and type of employment etc in the 
region. Therefore, by correlating the trip generations with socio-economic parameters, it 
is possible to develop mathematical models to predict the travel demand in terms of 
socio-economic variables. 
The objective of this study is to develop a family of trip generation models for the 
Colombo Metropolitan Region to estimate travel demand for work, education and other 
purposes by available motorized forms of transport. While the main focus is on 
estimation of bus passenger demand, another set of models is calibrated estimate 
aggregate demand for bus, rail, car, motor cycle and three - wheeler travel. Finally, a 
mode choice model is developed to estimate the variation of bus passenger modal share 
for work trips in terms of availability of rail and private vehicles. 
The regression facility available with SPSS V.10 software was used for the 
calibration of the models. Statistical testing methods such as R value, F-statistic, t-
statistic and residual analysis were used to identify the best predictor models. 
The calibrated traffic generation models can be used for estimating future trip 
generations in the Colombo Metropolitan Region. In addition, these models may be used 
for trip generation estimates for other geographic regions after validating for such 
regions. 
ACKNOWLEDGEMENT 
I express my sincere gratitude to Dr. Amal S. Kumarage, Head of Division of 
Transportation Engineering of University of Moratuwa for his guidance in completing 
this project. I am also grateful for the encouragement given by Dr. Saman Bandara, 
Senior Lecturer of Division of Transportation Engineering of University of Moratuwa. 
It is with great respect I extend my sincere thanks to Prof. Mrs. N. Ratnayake, 
Director of Graduate Studies of University of Moratuwa for arranging financial 
assistance and conducting research assessment programme. I wish to thank Dr. Kolitha 
Weerasekara, Senior Lecturer, Open University of Sri Lanka for his valuable comments. 
My sincere thanks to the fellow postgraduate researchers of the Division of 
Transportation Engineering, for the assistance and sharing their experience. 
Finally, I wish to thank the staff of the Division of Transportation Engineering for 
their support in conducting this research. 
TABLE OF CONTENTS 
TABLE OF CONTENTS i 
LIST OF TABLES vi 
LIST OF FIGURES viii 
LIST OF ABBREVIATIONS ix 
CHAPTER 1 1 
1.1 Background 1 
1.2 Organization of the Thesis 2 
CHAPTER 2 3 
2.1 Demand Estimation Model for Inter-District Passenger Travel (DEMIDEPT). 3 
2.1.1 Total Passenger Demand Estimation Model 4 
2.1.1.1 Variables Used 4 
2.1.1.2 Features of the DEMIDEPT Model 5 
2.1.1.3 Shortcomings of the DEMIDEPT Model 5 
2.1.2 Mode Choice Models 6 
2.1.2.1 Bus Vs Bus .' 6 
2.1.2.2 Bus Vs Rail 7 
2.1.3 Applications of DEMIDEPT 8 
2.2 Inter City Demand Estimation for Auto Travel from Link Counts 8 
2.2.1 Calibration of Demand Models 9 
2.2.2 Maximum Entropy Models 9 
2.2.3 Variables Used 9 
2.2.4 Features of the Models 10 
2.2.5 Shortcomings of the Models 10 
2.2.6 Applications of the Models 10 
2.3 Colombo Traffic Study (CTS/ TRANSPLAN -1) 11 
2.3.1 Land Use Based Trip End Model 11 
2.3.1.1 Variables Used 11 
2.3.1.2 Features of the Model 11 
2.3.1.3 Shortcomings of the Model 12 
2.3.2 Trip End Distribution Model 12 
2.3.2.1 Variables Used 12 
2.3.2.2 Features of the Model 12 
2.3.3 Route Choice Model 12 
2.3.3.1 Features of the Model 13 
2.3.3.2 Shortcoming of the Model 13 
2.3.4 Speed -Flow Model 13 
2.3.4.1 Variables Used 13 
2.3.4.2 Shortcomings of the Model 14 
2.3.5 Uses of TRANSPLAN - 1 Models 14 
2.4 Greater Colombo Traffic Model (GCTM/ TRANSPLAN - 2) 14 
2.4.1 Variables Used 14 
2.4.1.1 Socio-economic Variables 14 
2.4.1.2 Impredance Variables 15 
2.4.2 Features of the Models 15 
2.4.3 Shortcomings of the Models 15 
2.4.4 Applications of the Models 15 
2.5 A Mehod io Enhance the Performance of Synthetic Origin-Destination 
Trip Table Estimation Models 16 
CHAPTER 3 17 
3.1 Existing Models 17 
3.2 Selected Family of Models for Calibration 18 
3.3 Research Methodology 20 
2.4 Study Region 21 
CHAPTER 4 25 
4.1 Introduction 25 
4.1.1 Home Based Work Trips 27 
ii 
4 
4.1.2 Home Based Educational Trips 27 
4.1.3 Home Based Other Trips 27 
4.1.4 Non-home Based Trips 27 
4.2 Trip Generation Analysis 27 
4.2.1 Factors that Affect Trip Generation 28 
4.2.2 Models of Trip Generation 28 
4.3 Statistical Testing of Regression Analysis 31 
4.3.1 Coefficient of Determination (R 2) 31 
4.3.2 The Standard Error of Estimate (S) 31 
4.3.3 F-Test 31 
4.3.4 t-Statistic 32 
4.3.5 Statistical Significance 32 
4.3.6 Residual Analysis 33 
CHAPTER 5 34 
5.1 Data Used for Trip Generation Modelling 34 
5.1.1 Travel Data 34 
5.1.1.1 Sampling and Survey Errors 36 
5.1.2 Socio-Economic Data 37 
5.2 Data Preparation 37 
5.2.1 Household Trip Rates 37 
5.2.2 Socio-Economic Variables 38 
5.2.3 Categorization of Zonal Population by Employment Type 40 
5.2.4 Rail Station Density 41 
5.3 Preliminary Analysis 42 
5.3.1 Percentage Distribution of Trip Generations by Trip Purpose 43 
5.3.2 Analysis of the Variation of a Single Variable 43 
5.3.2.1 Home Based Work Trips per Household 44 
5.3.2.2 Home Based Educational Trips per Household 45 
5.3.2.3 Home Based Other Trips per Household 48 
5.3.2.4 Non-Home Based Trips per Household 50 
v 
iii 
5.3.2.5 Vehicle Ownership per Household 50 
5.3.2.6 Rail Station Density Distribution 54 
5.3.3 Analysis of Variation of Two variables 54 
CHAPTER 6 58 
6.1 Strategy 58 
6.2 Form of the Models 59 
6.3 Selection of the Base Variable 60 
6.3.1 Home Based Work Trips (HBWT) Vs Households in DSD (HHTOT) 60 
6.3.2 Home Based Work Trips (HBWT) Vs Population in DSD (POP) 61 
6.4 Home Based Work Trip Generation Model 62 
6.4.1 Analysis of Independent Variables 64 
6.4.1.1 Number of Households 64 
6.4.1.2 Private Vehicle Ownership 64 
6.4.1.3 Rail Station Density 68 
6.4.1.4 Percentage of School/University Population 68 
6.4.2 Prediction of Home Based Work Trips (Bus) for Validation DSD's in CMR69 
6.5 Home Based Education Trip (HBEDT) Generation Model 71 
6.5.1 Analysis of Independent Variables 72 
6.5.1.1 Number of Households 72 
6.5.1.2 Rail Station Density 73 
6.5.2 Other Observations 74 
6.5.3 Prediction of Home Based Educational Trips (Bus) for Validation DSD's in 
CMR 74 
6.6 Home Based Other Trip (HBOT) Generation Model 74 
6.6.1 Analysis of Independent Variables 76 
6.6.1.1 Number of Households 76 
6.6.1.2 Percentage of Unemployed Population 77 
6.6.2 Other Observations 77 
6.6.3 Prediction of Home Based Other Trips (Bus) for Validation DSD's in CMR78 
6.7 Discussion 78 
CHAPTER 7 
7.1 Introduction 
7.2 Home Based Total Work Trips 
7.3 Home Based Total Educational Trips 
7.4 Home Based Total Other Trips 
7.5 Discussion 
CHAPTER 8 
8.1 Introduction 
8.2 Modal Share Model For Home Based Work Trips (Bus) 
8.3 Sensitivity Analysis 
CHAPTER 9 
9.1 Conclusions 
9.2 Recommendations 
REFERENCES 
APPENDICES 
LIST OF TABLES 
Page 
Table 3.1: Available Demand Models 17 
Table 3.2: Selected Models for Calibration 18 
Table 3.3: DSD's within CMR 23 
Table 3.4: Calibration DSD's 24 
Table 3.5: Validation DSD's 24 
Table 5.1: Locations of Household Interviews and Sample Size 35 
Table 5.2: Travel Information Collection Duration at Different AGA's 36 
Table 5.3: Sample of Trip Generation Rate Calculation from Household Interviews 38 
Table 5.4: Annual Growth Rates Applied for Socio-economic Variables 39 
Table 5.5: Calculation of Socio-economic Variables for the Base Year = 1998 40 
Table 5.6: Rail Station Distribution Data for CMR 42 
Table 6.1: Statistical Comparison of Alternative Base Variables for Bus Model 61 
Table 6.2: Regression Statistics of HBWT Model 63 
Table 6.3: Regression Statistics of Independent Variables of HBWT Generation Model 
63 
Table 6.4: Predicted Trip Generation Rates for Validation DSD's 70 
Table 6.5: Regression Statistics of HBEDT Model 71 
Table 6.6: Regression Statistics of Independent Variables of HBEDT Generation 
Model 71 
Table 6.7: Regression Statistics of HBOT Model 75 
Table 6.8: Regression Statistics of Independent Variables of HBOT Generation Model 75 
vi 
Table 7.1: Regression Output for Total Home Based Work Trip (HBWTT) Generation 
Model 81 
Table 7.2: Regression Output for Total Home Based Educational Trip (HBEDTT) 
Generation Model 83 
Table 7.3: Regression Output for Total Home Based Other Trip (HBOTT) Generation 
Model 84 
Table 8.1: Estimated HBWT Values from Direct and Indirect Models 91 
Table 9.1: Summary of Demand Models 93 
LIST OF FIGURES 
Figure 2.1: Nested Mode Choice Model 6 
Figure 3.1: Alternative Methods for Bus Passenger Travel Demand Estimation 18 
Figure 3.2: Structure of the Family of Trip Generation Models Developed 20 
Figure 3.3: Colombo Metropolitan Region 22 
Figure 5.1: Percentage Distribution of Bus Trips by Purpose of Travel 43 
Figure 5.2: Home Based Work Trip Generation Rates per Household 46 
Figure 5.3: Home Based Educational Trip Generation Rates per Household 47 
Figure 5.4: Home Based Other Trip Generation Rates per Household 49 
Figure 5.5: Private Vehicle Ownership per Household 53 
Figure 5.6: Observed Rail Station Densities in CMR 55 
Figure 5.7: Modal Share of Bus Work Trips Vs Private Vehicle Ownership 56 
Figure 5.8: Modal Share of Bus Work Trips Vs Rail Station Density 57 
Figure 6.1: Bus Mode Based Work Trip Rate Vs Motor Cycle Ownership 66 
Figure 6.2: Bus Mode Based Work Trip Rate Vs Car Ownership 67 
Figure 8.1: Observed Variations in Work Trip Estimations Using Direct & Indirect 
Methods ....92 
viii 
LIST OF ABBREVIATIONS 
Abbreviation Description 
ADT Average Daily Traffic 
AGA Assistant Government Agent 
CMC Colombo Municipal Council 
CMR Colombo Metropolitan Region 
CMRSP Colombo Metropolitan Regional Structure Plan 
CTS Colombo Traffic Study 
CUTS-1 Colombo Urban Transport Study - Stage 1 
DEMIDEPT Demand Estimation Model for Inter District Passenger Travel 
DSD Divisional Secretariat Division 
EDUPC Percentage of Educational Population 
EXPO Exponential 
GCTM Greater Colombo Traffic Model 
HBEDT Home Based Educational Trips 
HBOT Home Based Other Trips 
HBWT Home Based Work Trips 
HH Household 
HHTOT Total Households per Zone 
MULTI Multiplicative 
NHBT Non Home Based Trips 
NTM National Traffic Model 
O-D Origin - Destination 
PV Private Vehicle 
PVHH Private Vehicles per Household 
STDEN Rail Station Density 
TRANSPLAN TRANSPLAN™ - University of Moratuwa 
UDA Urban Development Authority 
UNEMPC Percentage of Unemployed Population 
UoM University of Moratuwa 
ix