While transit exclusive lanes are statically introduced for an entire day (or for the peak period), the total number and distribution of trips vary from one hour to another during a day. In this study, a Transit Priority Network Design Problem (TPNDP) under variable demand has been formulated by a bi-level robust optimization approach. At the upper level, a non-linear mathematical program is adapted to find the optimum set of exclusive bus lanes at the network level such that the total travel time as well as its hourly variation is minimized and at the lower level, a four-step demand model is applied. The results show the importance of considering the hourly travel demand variation in designing a transit priority scheme and how such results can be compared with previous TPNDP models with a fixed demand. The proposed algorithm identifies the location of full time and part-time transit lanes in the network.

虽然整天（或高峰时段）会静态引入公交专用车道，但一天的行程总数和分布却从一个小时到另一个小时不等。在这项研究中，通过双层鲁棒优化方法制定了可变需求下的公交优先网络设计问题（TPNDP）。在较高级别，采用非线性数学程序来查找网络级别的专用公交专用道的最佳集合，从而使总行驶时间及其每小时变化最小化，而在较低级别，则分为四个步骤需求模型被应用。结果表明，在设计公交优先方案时考虑小时旅行需求变化的重要性，以及如何将这些结果与具有固定需求的以前的TPNDP模型进行比较。