Pedestrian crossing patterns in a gridded city, such as New York City, can take on a variety of forms. In the most simple case, a pedestrian may choose to walk a predetermined path in order to reach their destination. In other cases, the road that an individual decides to cross is dramatically influenced by which road has a cross signal, as well as the general willingness for an individual to wait for a do-not-cross signal to change. Through examining the variety of potential crossing patterns via an Individual Agent Model, we are able to quantify the optimal pedestrian crossing strategies in a gridded system.
- basicwalker.m: implements the Basic Model
- naivewalker.m: implements the Naive Model
- biasedwalker.m: implements the Biased Model
- varbiasedwalker.m: implements the Skewed Model
- trials.m: simulates many trials of the models and plots a distribution of the wait times
- optimal_bias.m: iterates over possible bias values to find the optimal one
- optimal_skew.m: iterates over possible skew values to find the optimal one
- bias_light_relationship.m: plots the relationship between bias and the length of a signal
- skew_light_relationship.m: plots the relationship between skew and the length of a signal
- aggregatepaths.m: plots the aggregate paths taken utilizing different strategies (requires modified code for walkers)
- trials_for_aggregate.m: creates aggregate (x,y) locations for each strategy
- basicwalker_mod.m, naive_walker_mod.m, biasedwalker_mod.m, varbiasedwalker_mod.m: implements same models, with outputs for aggregatepaths.m
All functions are written in MATLAB (R2014b)