TY - JOUR
T1 - Sample-based bayesian accident reconstruction
AU - Davis, Gary A.
PY - 2009/9
Y1 - 2009/9
N2 - In accident reconstruction, information derived from samples of crash tests can on occasion replace subjective prior information reflecting expert opinion. A method for accomplishing this, based on a straightforward application of Bayesian reasoning as used in statistics, is applied to three-accident scenarios: estimating critical speeds from yaw marks, estimating impact speed from crush, and estimating impact speed from pedestrian throw distance. The method is evaluated by comparing the estimates to measurements obtained in crash tests. In almost all cases the measured speeds were captured by the posterior 95% credible intervals.
AB - In accident reconstruction, information derived from samples of crash tests can on occasion replace subjective prior information reflecting expert opinion. A method for accomplishing this, based on a straightforward application of Bayesian reasoning as used in statistics, is applied to three-accident scenarios: estimating critical speeds from yaw marks, estimating impact speed from crush, and estimating impact speed from pedestrian throw distance. The method is evaluated by comparing the estimates to measurements obtained in crash tests. In almost all cases the measured speeds were captured by the posterior 95% credible intervals.
KW - Accident reconstruction
KW - Bayesian inference
KW - Markov chain Monte Carlo
UR - http://www.scopus.com/inward/record.url?scp=79959507650&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=79959507650&partnerID=8YFLogxK
U2 - 10.1080/19439960903149579
DO - 10.1080/19439960903149579
M3 - Article
AN - SCOPUS:79959507650
SN - 1943-9962
VL - 1
SP - 181
EP - 189
JO - Journal of Transportation Safety and Security
JF - Journal of Transportation Safety and Security
IS - 3
ER -
