PARAMETER OPTIMIZATION METHOD FOR EXPRESSWAY TRAVEL TIME ESTIMATION MODEL

摘要

Disclosed is a parameter optimization method for an expressway travel time estimation model. The method comprises: constructing, based on a road section travel time estimation model, a route travel time estimation model including a plurality of road sections; optimizing model parameters l, m and α by means of a genetic algorithm; defining a genetic algorithm fitness function, wherein Wj is a weight of an error index MAj, j is a time interval of travel time estimation, a value range thereof is [2, K], and an optimization target of the genetic algorithm is that the fitness function is the minimum; calculating the weight Wj by means of an M-H algorithm; outputting optimal parameter combinations [l, m, α] in a free flow state and in a congested state by means of the genetic algorithm; and then establishing an optimized route travel time estimation model. In the method, the parameters of a travel time estimation model are automatically optimized, such that the sensitivity of the model in different traffic operation states such as a congested state and an unobstructed state is improved, the travel time estimation precision in a congested situation is significantly improved, and the reliable and stable performance of the model is guaranteed.