Voir la source de Surrogate Measures of Safety/Minimal Road User Interaction Simulation

Master's thesis by [[Utilisateur:LionelNebotJanvier]].

==Objectives==
General objective: to generate simple road user trajectories from theoretical or empirical headway distributions and compute distributions of surrogate measures of safety (SMoS) like time to collision and post-encroachment time. 

Specific objectives:
* to study the impacts on TTC and PET distributions of
** parameters of the scene and "instrumentation": the size of the area of interest, analog to the camera field of view) 
** road characteristics: type of traffic control devices, in particular traffic lights compared to yield signs (minor/major road)
** traffic characteristics: volume, headway distributions, etc.

Case studies: pedestrian crossing, intersection of vehicles with traffic lights, stop signs or yield (roundabout?)

==Methodology==
* Collect and characterize road user headway distributions
* Develop the model of the scene and road users, and the tools to save/load them: how many parameters, extensible model?
* Generate trajectories and compute SMoS

==Choices==
* Use of curvilinear coordinate system: interactions can occur within each link (rear end) and only at nodes for road users from different streams
* If movement at constant speed, there is a need to include perception-reaction time for the road user without priority to react (is there?)

To be decided:
* road user volume (instead of centroid/circle representation)
* minimal graphical representation
* implement more than one lane: probably no (not necessary)

==Resources==
* General resources [[Surrogate_Measures_of_Safety#Resources]]
* Estimating the severity of safety related behaviour, Åse Svensson and Christer Hydén https://doi.org/10.1016/j.aap.2005.10.009