Other ongoing research activities

Transcription

1 Go look at the posters during lunch! Other ongoing research activities Pre Catherine Morency, ing., Ph.D. Head of Chaire Mobilité Hubert Verreault, ing. jr., M.Sc.A. 5 e Annual Meeting, Polytechnique Montréal, May 26th 2015

2 Outline Recent topics / in development Bridge choice model Longitudinal survey Other ongoing projects Long-distance travel intercity car Taxi P P

3 BRIDGE choice model (1) Paper: Application of an Independent Availability Logit Model (IAL) for Route Choice Modelling: Considering Bridge Choice as a Key Determinant of Selected Routes for Commuting in Montreal, Journal of Choice Modelling, (Dr. Khandker Nurul Habib et al. 2014) Presentation at AQTR 2015 Quand le pont dicte le trajet!» Objectives : Contribute to : Validate and value data collected during OD surveys regarding the selected bridge (data that have been collected for quite some time now) Better understand route choice behaviours Test new modelling approaches demonstrate possibilities Example of an assessment tool (scenario analysis) regarding bridges (lane closure, roadwork)

4 Assumption BRIDGE choice model (2) Bridge choice has an important impact on the route choice between an O-D Test OD Car driver trips declared bridge = bridge choice model (proxy for route choice) For each OD pair: Shortest path O-P i -D Travel time % superposition of route between each pair of route

5 Exemple : 3 alternatives Which alternatives are taken into account by the traveler before he chooses? THE MODEL What is the probability of each alternative? Set of alternatives Pont 1 Possible alternatives (j) Pont 2 Bridge choice Pont 3 Pont 1 / Pont 2 Pont 1 / Pont 3 Pont 1 Pont 2 Pont 2 / Pont 3 Pont 1 / Pont 2 / Pont 3 Pont 3 Probability for each choice set (EA) Probability of each alternative within a choice set Probability of each alternative Probability P EA 5e Colloque Annuel de la Chaire PMobilité J EA P j = P EA P j EA

6 BRIDGE choice model(3) Results available in the paper and presentation Advanced econometric model can assist in the processins of declared data on bridge choice - to describe partial route choice : Brige attributes Trip and alternatives attributes Traveller attributes Demographic variables + proxy for neighborhood + relevant vs various scenarios (income vs value of time for instance = toll) Limits! Congestion not taken into account for now + next step = OD 2013

7 Adding value to continuous OD survey data (1) Experimental survey conducted by AMT from January 2009 to decembre Objective: add value to these data + put them in relation with OD 2008 and 2013 Test (longitudinal approach): Analysis and modelling of people-based indicators Analysis and modelling of mobility indicators (key figures per day/week/month) Inclusion of exogeneous variables (weather, fuel price, )

8 Adding value to continuous OD survey data (2) Person-based approach Outcome variable Demography (people, household) No weighing is required, behavioural model IdPers Trip rate Year Month Week Day Age Gender Occupati on Driving licence HH size Weather Fuel price Temporal circumstances Exogeneous variables We are currently starting the modelling work on these data

9 Adding value to continuous OD survey data (3) Person-based approach Ordinal LOGIT STUDY trips per day Depl_Étude Rapport de cote Std. Err. z P>z log_age dy dy dy dy2009 (omitted) ddow ddow ddow ddow ddow1 (omitted) dweek dweek37 (omitted) dmen dwom (omitted) ddriv_lic E+01 0 docc_sta E+02 0 dregdomi E dregdomi E dhhz E dhhz dhhz E+00 0 dhhz4 (omitted) dhco E dhco E dhco E Reduction in 2012 Lower on fridays

10 Rapport de cote Adding value to continuous OD survey data (4) Person-based approach Odds ratio of WEEK variable Semaine de l'année Rapport de cote School break Reference week = 37 (mid-september) End CEGEP End primary + secondary schools Start of school year

11 Adding value to continuous OD survey data (5) Key figures: example, / week Outcome variable Sample composition No weighing, variables controling for sample composition in the model Id week Trip rate Modal share Year Month Week % by region % population segment Weather Fuel price Temporal circumstances Exogeneous variables

12 Taux de déplacement Prix de l'essence Adding value to continuous OD survey data (6) Key figures: example, /week Car driver trips per person Année, semaine prix_ess deplmac

13 Adding value to continuous OD survey data (6) Various modelling experimentations to be conducted (variables, Model formulation) collaboration U of T Challenges (among others): Sample size No a priori weighing! Sample composition over time (space, population segment) Variabilité of behaviours vs measures vs long term trends

14 Long-distance travel (intercity bus) P At last! This project is starting 4.1 Literature review + state of practice 4.2 Integrate database on supply and demand for intercity bus in Quebec Supply Demand 4.6 Method for the estimation of the potential market of buses 4.5 Tool for the spatial-temporal analysis of supply and demand 4.7 Prototype: long-distance travel survey 4.8 Measure of the contributions to sustainable development

15 1 Taxi: building knowledge on supply Literature review + state of practice in Montreal and demand 2 Integrated systems for GPS + transactions for multiple service providers GPS data P Partners: Bureau du taxi de Montréal Agence de revenus Intermédiaire en service 3 Performance indicators + method for the estimation 5 4 Current and potential market of taxi trips in Montreal Tool for the recurrent diagnostic 6 OD survey among customers + link with regional surveys Contribution of taxi to sustainable development

16 Other ongoing projects Congestion monitoring tool «Code de la rue» and how it can be applied in Quebec GTFS data finding new applications Dynamic ridesharing- NETLIFT Road sharing (indicator) and link with safety Young people and the car (France-Québec)

17 Lunch! Please visit the POSTERS! And be back for 1h30 pm!