Search

The calibration of urban traffic microsimulation models is addressed in this paper, focusing on the multivariate distribution of traffic features. This allows to take into account day-to-day variability and possibly complex statistical dependencies, enabling robust validation and the computation of more accurate statistics. Variability in traffic models can be obtained, in principle, by suitably tuning origin-destination demand flows, but this becomes problematic for microsimulation models when complex multivariate distributions are considered, due to excessive dimensionality and nonlinearity. Thus, here we investigate how complex multivariate behavior can be achieved by suitably varying only a subset of the parameters ruling the microsimulation dynamics, without imposing any distribution on the demand flows. To this purpose, rather than looking for fixed values of the parameters, we exploit distribution models having the structure of mixtures of uniforms, through which the simulator parameter values can be randomly sampled, thus replacing the simulator inner randomization mechanism. We formalize the optimization of the mixture parameters through a maximum mean discrepancy principle involving the simulator dynamics, which leads to a large-dimensional non-differentiable problem that we solve through a method based on cross-entropy. Preliminary experiments concerning the application of the proposed methodology to the SUMO simulator show how it is possible to capture quite complex multivariate distributions of target flows varying only four parameters. [ABSTRACT FROM AUTHOR]

In this paper, a general method for the numerical solution of maximum-likelihood estimation (MLE) problems is presented; it adopts the deterministic learning (DL) approach to find close approximations to ML estimator functions for the unknown parameters of any given density. The method relies on the choice of a proper neural network and on the deterministic generation of samples of observations of the likelihood function, thus avoiding the problem of generating samples with the unknown density. Under mild assumptions, consistency and convergence with favorable rates to the true ML estimator function can be proved. Simulation results are provided to show the good behavior of the algorithm compared to the corresponding exact solutions. Index Terms--Deterministic learning (DL), discrepancy, maximum-likelihood estimation (MLE), variation.

A methodology to design state estimators for a class of nonlinear continuous-time dynamic systems that is based on neural networks and nonlinear programming is proposed. The estimator has the structure of a Luenberger observer with a linear gain and a parameterized (in general, nonlinear) function, whose argument is an innovation term representing the difference between the current measurement and its prediction. The problem of the estimator design consists in finding the values of the gain and of the parameters that guarantee the asymptotic stability of the estimation error. Toward this end, if a neural network is used to take on this function, the parameters (i.e., the neural weights) are chosen, together with the gain, by constraining the derivative of a quadratic Lyapunov function for the estimation error to be negative definite on a given compact set. It is proved that it is sufficient to impose the negative definiteness of such a derivative only on a suitably dense grid of sampling points. The gain is determined by solving a Lyapunov equation. The neural weights are searched for via nonlinear programming by minimizing a cost penalizing grid-point constraints that are not satisfied. Techniques based on low-discrepancy sequences are applied to deal with a small number of sampling points, and, hence, to reduce the computational burden required to optimize the parameters. Numerical results are reported and comparisons with those obtained by the extended Kalman filter are made. Index Terms--Feedforward neural networks, Lyapunov function, offline optimization, penalty function, quasi-random sequences, state observer.

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.ejor.2005.01.022 Byline: Cristiano Cervellera (a), Victoria C.P. Chen (b), Aihong Wen (b) Keywords: Dynamic programming; Large-scale optimization; Applied probability; Neural networks; Natural resources Abstract: A numerical solution to a 30-dimensional water reservoir network optimization problem, based on stochastic dynamic programming, is presented. In such problems the amount of water to be released from each reservoir is chosen to minimize a nonlinear cost (or maximize benefit) function while satisfying proper constraints. Experimental results show how dimensionality issues, given by the large number of basins and realistic modeling of the stochastic inflows, can be mitigated by employing neural approximators for the value functions, and efficient discretizations of the state space, such as orthogonal arrays, Latin hypercube designs and low-discrepancy sequences. Author Affiliation: (a) Institute of Intelligent Systems for Automation, ISSIA-CNR National Research Council of Italy, Genova Branch, Via De Marini 6, Genova 16149, Italy (b) Department of Industrial and Manufacturing Systems Engineering, The University of Texas at Arlington, Campus Box 19017, Arlington, TX 76019-0017, United States