基于成本迭代与需求迭代的四阶段交通模型双重反馈算法研究

Study on Four step Model’s Dual Feedback Algorithm Based on Cost Iteration and Demand Iteration

中文关键词:四阶段模型  需求预测  反馈算法  成本迭代  需求迭代

英文关键词:four step model  demand forecast  feedback algorithm  cost iteration  demand iteration

中文摘要:

      四阶段交通模型作为最流行的交通需求预测模型理论，其计算结果存在不一致性的缺陷，反馈算法被引入模型结构中来解决不一致性的问题。研究提出了成本迭代与需求迭代的双重迭代反馈算法，并使用案例比较了它与既有反馈算法的差异。研究结果表明:成本迭代＋需求迭代的双重迭代模型结构能够在不影响预测结果的前提下减少迭代次数和系统运行时间，收敛过程更加稳定且能够适用于不同规格的交通网络。

英文摘要:

      The four step model is the most popular traffic demand forecast model theory, but its result contains an internal inconsistency. Feedback algorithm is introduced to maintain the data consistency of the various steps of the model structure. This study proposes dual iteration feedback algorithm based on cost iteration and demand iteration, then compares it with now used feedback algorithms through different cases. The result shows the new model structure can keep the model result and reduce number of iterations and running time effectively, its convergence process is more stable than current model structures and can be applicable for different size networks.