Development of a Time invariant Stochastic Model of a Transport Mission

Rapport, 2005

In this report a continuous state space markov process is identified from drive data collected in central Gothenburg. The drive behavior is modeled as a time invariant markov process with acceleration and velocity as the markov states. The acceleration is modeled as beta distributed with the beta parameters dependent on previous acceleration and velocity. The model is verified by comparing simulated velocities and accelerations with the measurements. The simulations show that the model is able to reproduce most of the behavior exhibited in the measurements. It is therefore concluded that the beta distribution is a good modeling choice for a markov process that describes the time invariant distribution of acceleration and velocity.