A SENSITIVITY-BASED SOLUTION ALGORITHM FOR THE NETWORK MODEL OF URBAN TAXI SERVICES

摘要

A sensitivity-based algorithm has been proposed to solve the taxi model with elastic demand in a congested road network. In this algorithm, the Jacobian matrix of the SLNE is expressed explicitly as a function of the solution from the CNEM results, and is renewed at each SLNE iteration to maintain the best quality of likely descent direction. It has been found from an extensive computational experiment that the sensitivity-based algorithm is much more efficient than the quasi-Newtonian algorithm proposed in Wong et al. (2001) in which an initial Jacobian matrix needs to be computed by means of a computationally intensive finite difference method at the first iteration.