Optimal investment planning for improving highway network performance in the context of tropical cyclones

Pre-disaster investment planning for improving highway network performance is one of key measures to reduce disaster adverse impact and risk for transport agency. This paper seeks to the optimal investment decision to retrofit highway network against tropical cyclones. Monte-Carlo simulation method was used to generate 100-years tropical cyclone tracks in order to capture the pattern of tropical cyclones in Hainan province. Based on the generated tropical cyclone scenarios, a bi-objectives stochastic model is developed by considering investment cost and highway network performance. The proposed model is a NP-hard problem, whose exact solution cannot be found out in polynomial time. Non-dominated sorting genetic algorithm gives a promising perspective for searching the optimal solutions. Hainan province in China severely suffers from tropical cyclones, where there emerges an urgency to upgrade the highway network performance in tropical cyclones. We took Hainan province as an example to illustrate the efficiency and effectiveness of the proposed model and algorithm.