Urban development usually deteriorates the transportation system. For sustainable urban development, policymakers often face the challenging problem of how to optimally allocate overall land use quotas across a number of residential locations according to the performance of the transportation system. This is a kind of Stackelberg competition, where policymakers make land use decisions and travelers make behavioral responses. A novel bi-level model is formulated to solve this problem. The upper-level model minimizes the total system travel time by land use allocation, while at the lower level are sequential models with feedback for transportation system equilibrium. The Dirichlet allocation algorithm, a simulation-based heuristic algorithm, is designed to solve this bi-level model. A simulation experiment using the Nguyen–Dupuis network is then used to verify the proposed model and algorithm. The results from the simulation experiment demonstrate that not only are the model and algorithm operational but that they also provide an effective tool for policymakers to plan for land use development.

中文翻译：

---

交通导向的土地利用开发空间配置：一种基于模拟的优化方法

城市发展通常会使交通系统恶化。对于可持续的城市发展，政策制定者经常面临如何根据交通系统的性能在多个住宅区优化分配总体土地使用配额的挑战性问题。这是一种斯塔克尔伯格竞赛，政策制定者做出土地使用决策，旅行者做出行为反应。一个新的双层模型被制定来解决这个问题。上层模型通过土地利用分配使系统总旅行时间最小化，而下层模型是具有交通系统平衡反馈的序列模型。Dirichlet 分配算法是一种基于模拟的启发式算法，旨在解决这种双层模型。然后使用 Nguyen-Dupuis 网络的模拟实验来验证所提出的模型和算法。模拟实验的结果表明，模型和算法不仅是可操作的，而且还为政策制定者规划土地利用开发提供了有效工具。