Scholarlink

<p>We present a methodology for calibrating doubly constrained gravity models. Three types of deterrence functions are examined: exponential, power, and combined. The latter includes two calibration parameters, whereas the others contain only one. The parameters are identified so that the model reproduces the observed trip length distribution with high accuracy. We test four optimization methods: Nelder–Mead method, genetic algorithm, particle swarm optimization, and covariance matrix adaptation evolution strategy and analyze three forms of the objective function. The proposed methodology is applied to data from the city of Ekaterinburg. As a result, we provide recommendations regarding the form of the deterrence function, optimization method, and objective function. Bibliography: 17 titles. Illustrations: 5 figures.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

CALIBRATION OF GRAVITY MODELS

  • 期刊论文

  • 发表日期：

    2025年12月1日

查看全文
  • Valentina Kurtc,
  • Andrey Prokhorov

摘要

We present a methodology for calibrating doubly constrained gravity models. Three types of deterrence functions are examined: exponential, power, and combined. The latter includes two calibration parameters, whereas the others contain only one. The parameters are identified so that the model reproduces the observed trip length distribution with high accuracy. We test four optimization methods: Nelder–Mead method, genetic algorithm, particle swarm optimization, and covariance matrix adaptation evolution strategy and analyze three forms of the objective function. The proposed methodology is applied to data from the city of Ekaterinburg. As a result, we provide recommendations regarding the form of the deterrence function, optimization method, and objective function. Bibliography: 17 titles. Illustrations: 5 figures.