Calibration of Microsimulation with Heuristic Optimization Methods

Abstract

Model calibration is a crucial step in building a reliable microscopic traffic simulation application, because it serves as the additional check to ensure the model parameters accurately reflect the local driving environment, such that decisions made based on these results would not be misinformed decisions. Because of its stochastic nature and complexity, the calibration problem, usually formulated as an optimization problem, is often solved using heuristic methods. To-date, calibration is still a timeconsuming task because many of the adopted methods require many simulation runs in search of an optimal solution. Moreover, many aspects of the calibration problem are not fully understood and need further investigation. In this study, we develop another heuristics calibration algorithm based on the simultaneous perturbation stochastic approximation (SPSA) scheme, and applied it to calibration several networks coded in Paramics. Our results indicate that the new heuristic algorithm can reach the same level of accuracy with considerably less iterations and CPU time than other heuristic algorithms such as the genetic algorithm (GA) and the trial-and-error iterative adjustment (IA) algorithm. Applications of all three heuristic methods in a northern California network also reveal that some model parameters affect the simulation results more significantly than others. These findings can help modelers better choose calibration methods and fine tune key parameters.

Calibration of Vissim Model for Multilane Highways Using Speed Flow Curves

“…CC0 (standstill distance) and CC1 (Time Headway in seconds) as these two parameters are the most effective parameters for calibration, the values suggested by Shukla and Chandra. Jingtao et al [10] studied calibration of micro simulation with heuristic optimization methods and proposed a new heuristic calibration algorithm which is based on simultaneous perturbation stochastic approximation (SPSA) scheme. The results obtained using this technique have the same level of accuracy with considerably less iteration and less time as compared to other heuristic algorithms like Genetic algorithm (GA) and the trial and error iterative adjustment (IA) algorithm.…”

Holistic Calibration of Microscopic Traffic Flow Models: Methodology and Real World Application Studies

“…Genetic Algorithms were used for the calibration of global and local capacity and occupancy parameters [20,29]. A sequential approach was used to update global and local parameters.…”

“…In addition, the SPSA algorithm has been used to calibrate and optimize various transportation applications [13,19,20].…”

Parameter Calibration Method of Microscopic Traffic Flow Simulation Models based on Orthogonal Genetic Algorithm

“…Genetic algorithm (GA) has become the most common used calibration algorithm for input parameters of the simulations [5,[10][11][12][13], since Cheu et al, firstly used GA calibrating FRESIM model [14]. Other intelligent algorithms are also used in the calibration of traffic simulation, e.g., perturbation stochastic approximation (SPSA) scheme [15], particle swarm optimization (PSO) [16], and neural network approach [4]. These methods automate the calibration process to a certain degree and it was generally reported that they improve simulation performance over the default model parameter values.…”