Dear Researchers,
I need to calculate the probability of the vehicles having some kind of incidents/accidents based on the intersection characteristics (grade, signage, median, etc.). It will answer the question that, for example, how much the probability of head-on crashes would change if we add raised median to the intersection.
However, I was wondering if there is any method/model to be used that relates this probability to actual frequency. For instance, back to my previous example and given the probability of head-on crashes, we can determine 5 head-on crashes will occur on a given intersection.
Thank you!
I think the answer is simple: usually, the frequency is a good estimate of the probability, and the probability is also used as an estimate of a frequency. If we say that the probability to miss the bus means that in many such similar occasions we expect to miss the bus with a frequency equal to this probability.
Dear Mahdi,
To the best of my knowledge, i've seen such cases (in analogy) to mobile nodes in some sort of network. Well, say you have N mobile nodes (or N cars) moving in a certain cell (or Area), then the probability that two nodes (or two cars) intersect (or crash) with each other follows a Poisson distribution of frequency (or average rate). say lambda. After that, you could form a Markov chain of events.
I'm really not sure though how the Markov chain is done. But what i'm sure of is that a simple case to model such a situation is a Poisson distribution.
addting to Ahmad: ... or a binomial distribution, when the maximum possible number of events is known.
Dear Mahdi, I suppose that crash prediction models are what you are looking for. They relate crash frequency to various explanatory variables, which is exposure (traffic volume) and geometrical characteristics, such as median, etc. Since crash frequency is a count variable with non-normal distribution, they are usually developed using generalized linear models with negative binomial distribution. This is also what I am using; however for the work of real guru look for example here - https://www.researchgate.net/publication/239438865_Statistical_Road_Safety_Modeling
Article Statistical Road Safety Modeling
Dear Mahdi,
All that you need is the sample size that the probabilities have been calculated based on. If you could find out the sample size, you can turn it to frequency just by multiplying the N and The P. Otherwise, you can calculate frequencies for your population of interest in the same way.
Dear All,
Thank you so much for your responses! They actually gave me a lot of clues! Let me go through your posts and try to figure out the points, I will definitely come up with other questions and post them here!
Again, thank you for helping me on this topic.
Mahdi
- Badges
- Science topic
- Similar topics
- Geoscience
- Asia
- India