Transport OP_lecture 1.pdf

CVEN4003W Transport Operation and Planning Lecture 1 Dr. Meisam Gordan School of Civil Engineering University College Dublin What is Transport Modelling? 2  What is Transport Modelling? The study of the behaviour of individuals, associated with: – the move...

CVEN4003W Transport Operation and Planning Lecture 1 Dr. Meisam Gordan School of Civil Engineering University College Dublin What is Transport Modelling? 2  What is Transport Modelling? The study of the behaviour of individuals, associated with: – the movement of themselves (passenger transport)  What is Transport Modelling? The study of the behaviour of individuals, associated with: – the movement of themselves (passenger transport) – Commodities (freight transport)  What is Transport Modelling? The study of the behaviour of individuals, associated with: – the movement of themselves (passenger transport) – Commodities (freight transport) – Information via telecommunications  What is Transport Modelling? The study of the behaviour of individuals, associated with: – the movement of themselves (passenger transport) – Commodities (freight transport) – Information via telecommunications The models are abstractions of the real world that are based on some mathematical equations but most importantly physical intuition.  What is a Transport Model? A representation of a transport system, built to simulate existing or forecast future traffic conditions in order to inform a decision-making process. Simulation - attempting to replicate a real life, observed situation Forecasting - attempting to predict a situation (usually in the future) 7  What is a Transport Model? – Example A transport model is a set of mathematical equations or rules that tries to describe or replicate a physical process. In our case, the physical process that we are studying is evolution of traffic congestion over time and space. 8  Why do we need models? Process analysis and understanding Planning (forecasting) – Introduction of a new mode (multimode) Evaluate Evaluate&& Overview Overview Define Definethe the Implement Implement Formulate Formulate Evaluate Evaluate Start Start select select data data problem problem alternatives goalalternatives goal outcome outcome alternatives alternatives – Modification/ extension of infrastructure Generate Operations (control) Operationalize Operationalize Generate Compute Computethethe goal goal alternatives alternatives optimal optimaldecision decision – Design of model-based Identify Identify control Abstract intostrategies Abstractinto decision decisioncriteria criteria Model Model – Estimation/ prediction models – Testing the control Weight performance Weight criteria criteria No No Does DoesDecide Decide solution solution subjectively subjectively Traffic Simulation exist? exist? Yes Yes Gain Gain Predict Predict Rate Rateeach each insight insight performance performance alternative alternative  Why do we need models? Process analysis and understanding Planning (forecasting) – Introduction of a new mode (multimode) – Modification/ extension of infrastructure Operations (control) – Design of model-based control strategies – Estimation/ prediction models – Testing the control performance Traffic Simulation  Why do we need models? – Example Imagine we have a transport system, and we want to introduce a new mode, for example, a new metro line. We can use a model to assess the impact of this change in the infrastructure. Also, if you want to modify or extend your infrastructure, Example: The for example, to build travel a Time Travel new highway, again you times and can use ofmodelsMode costs to do the planningCostof(Euro) (minutes) the new system different and assess theCarperformance transportation 30 of the10new system. modes before and after Bus 45 5 the Another use of oftraffic introduction a models Metro is for 20operations3 and traffic new control. metro line: So, we can use analytical mathematical models to design model-based control strategies, to build estimation and prediction algorithms and to test the performance of our control algorithms.  Why do we need models? – Example Also, if we want to modify or extend our infrastructure, for example, to build a new highway, again we can use models to do the planning of the new system and assess the performance of the new system. Another use of traffic models is for operations and traffic control. So, we can use analytical mathematical models to design model-based control strategies, to build estimation and prediction algorithms and to test the performance of our control algorithms. Imagine you have a transport system, and you want to introduce a new mode, for example, a new metro line or a new bus line. You can use a model to assess the impact of this change in the infrastructure.  Why do we need models? – Example Another use of traffic models is for operations and traffic control. So, we can use analytical mathematical models: → to design model-based control strategies, → to build estimation and prediction algorithms and, → to test the performance of our control algorithms. Imagine you have a transport system, and you want to introduce a new mode, for example, a new metro line or a new bus line. You can use a model to assess the impact of this to build a new highway, again you can use models to do the planning of the new system and assess the performance of the new system. Representing reality 14  What is Transport Modelling – the simple answer An analytical tool Replicate choice behaviour and travel patterns using mathematical equations and empirical assumptions 15 Multi-disciplinary  Essentials of a transport model A spatial representation of the supply of transport and of the demand to travel – A– matrix of trips A matrix between of trips betweenorigins andand origins destinations destinationsin the model, in the model, assigned to the assigned transport to the network transport networkandandallocated to the allocated modes to the modes represented, based represented, on on based users minimising users their minimising disutility their of travel disutility of travel – A– process A process which which allows allows for for testing testing options, options, with with accessibility accessibility measured measured through through travel travel times times andand costs, costs, thethe market market for for those those changes changes andand a new a new equilibrium equilibrium between between demand demand andand supply supply – A transport model is invariably augmented by a forecasting – A transport model is invariably augmented by a forecasting model – future year demand drivers and responses to the model – future year demand drivers and responses to the changes in future years changes in future years – Output – changes in the quantity and quality of travel – Output – changes in the quantity and quality of travel Data requirements for building a model SURVEY DATA DEMOGRAPHIC DATA Automatic Traffic Counts (ATC) Census information (by area) Roadside Interviews (RSI) Household + Jobs data (by Automatic Number Plate area) Recognition (ANPR) Sat-Nav/GPS data Car Ownership statistics SITE SPECIFIC DATA FORECASTING DATA GIS Layers Traffic Growth Forecasts Google Earth/Streetview Fleet Composition forecasts Latest on-site observations Traffic Emission forecasts Local knowledge Development Plans  Data The data used in transport modelling is compiled from samples of the population Sampling Methods Simple Random Sampling This method involves associating an identifier (number) to each unit in the population and then selecting these numbers at random to obtain the sample - Problem: Far too large a sample would be required to ensure sufficient data is collected on minority groups Stratified Random Sampling In this method, the population is subdivided into homogeneous strata and then random samples are taken from each of these groups Sampling Error and Sampling Bias There are two types of errors that can be introduced in sampling Error generated due to the fact that the sample is only a proportion of the population. The sampling error is therefore a function of the sample size Sampling bias is caused by mistakes made either 1. When defining the population of interest or 2. When selecting the sample method  The sample size can be estimated by using the Equation below (Dillman, 2000) 𝑁𝑝𝑝 𝑝𝑝 (1 − 𝑝𝑝) 𝑁𝑠 = 𝐵 (𝑁𝑝𝑝 − 1)( )2 + (𝑝𝑝)(1 − 𝑝𝑝) 𝐶 the confidence level Where: Ns = sample size required for the desired level of precision Npp = size of population pp = proportion of the population expected to choose one of the two response categories B = acceptable amount of sample error C = Z statistic associated with the confidence level Errors in Modelling and forecasting The ideal requirement is to find a combination of model complexity and data accuracy which best fits the required forecasting precision and study budget Measurement Errors Survey questions badly interpreted, answered badly coding errors, etc., can cause measurement errors Sampling Errors These errors are due to the representation of the population by finite data sets. One way to reduce the sample error is to increase the sample size, however this can be costly Aggregation Errors Typically, in models the forecasting is done for groups of individuals but the data is compiled on the basis of the responses of the individuals  Basic Data-Collection Methods Selection of the most appropriate data collection methods will depend significantly on the type of models that will be used in the study Will define what type of data is needed and therefore what data collection methods are more appropriate Very important to achieve a balanced study (in terms of its various stages), i.e. largest part of the study budget (and time) is spent in data collection, analysis and validation (Boyce et al., 1970)  Study Horizon If the design year is too close, there may not be sufficient time to conduct the modelling study. Will imply need to use a particular analysis tool, requiring data of a certain type. In strategic transport studies, usual horizon is 20 or more years into future. Allows more time for most types of analytical tools