I need to simulate for an academical project how the traffic fluxes (input/output with respect to a monitored area, measured in number of cars) of a city area evolves in correspondence of an event (i.e. the opening of a restricted traffic area to decongest the traffic). I have some simulated sensors that provide the data: I was thinking to use a combination of a fuzzy system (to assign a membership function to each type of data, e.g. PM10 value and CO2 value) and a markov process: I would need to modify the probability to decrement the number of car in the monitored area (simulating that a car is going out the congested area, towards the new opened area) basing on decisions made by means of a fuzzy system. So my questions are:

  • It is a good way to interpret the problem or there are better ideas that I have not taken into account yet?
  • How to implement such a combination of markov chain and fuzzy systems in matlab?



1 Answer 1


I don't really get why would you mix Fuzziness and Probabilities. HMMs already can give you probabilities without the need of adding Fuzzy systems into the mix.

I would just do a random walk with probabilities of transitions defined by the state of the lights.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.