In this paper, we determine a welfare-optimal investment strategy for bicycle networks while considering the joint impact of travel time savings and induced demand throughout the investment horizon. The paper extends an expansion strategy with fixed demand recently published in Paulsen and Rich (2023). Accommodating induced demand requires that we can express how expansions of the network affect demand through changes in travel time, and also how demand contributes to the net present value function due to accumulated health and travel time savings. It is shown how the net present value function can be approximated through approximations of demand and consumer surplus functions enabling optimisation on a large scale. This process is formulated as a sequence of binary mathematical programs spanning the entire period. We test the approach by applying it to a large-scale network of Cycle Superhighways for Greater Copenhagen. It is demonstrated that the optimised infrastructure investment plan renders benefit–cost ratios exceeding 10 and that accounting for demand effects in the optimisation increases the societal net present value by more than 15% while also changing the geographical structure of the invested network. The paper emphasises the significance of approaching infrastructure investment strategies from a dynamic perspective instead of relying on static evaluations, which is the predominant practice in practical planning.

中文翻译：

---

福利最优自行车网络扩展与诱导需求


在本文中，我们确定了自行车网络的福利最优投资策略，同时考虑了整个投资范围内旅行时间节省和诱导需求的共同影响。该论文扩展了最近发表在 Paulsen 和 Rich （2023） 上的固定需求扩张战略。适应诱导需求要求我们能够表达网络的扩展如何通过旅行时间的变化影响需求，以及需求如何由于累积的健康和旅行时间节省而对净现值函数做出贡献。它显示了如何通过需求和消费者剩余函数的近似来近似净现值函数，从而实现大规模优化。这个过程被表述为跨越整个时期的二进制数学程序序列。我们通过将其应用于大哥本哈根的大规模自行车高速公路网络来测试该方法。结果表明，优化的基础设施投资计划使收益成本比超过 10，并且在优化中考虑需求效应使社会净现值增加 15% 以上，同时也改变了投资网络的地理结构。本文强调了从动态角度处理基础设施投资策略的重要性，而不是依赖静态评估，而静态评估是实际规划中的主要做法。