Optimization of Road Network Recovery Decisions Considering Road Section Recovery Differences

Abstract

Existing studies on road network recovery decision have ignored the impact of the differences in recovery speed and recovery degree of different road sections on the recovery performance of the road network. To address this problem, road network connectivity index based on section impedance tolerance was first constructed to evaluate road network performance under partial recovery of road section capacity. Then, a bi-level optimization model for emergency recovery decisions was constructed with the weighted road network performance resilience and recovery speed resilience as optimization objectives. When the optimal set and recovery time sequence of the road sections to be repaired are determined, the recovery degree and speed of the road sections to be restored are obtained through resource allocation and budget allocation at the road section level. Finally, based on the traditional parallel machine scheduling problem genetic algorithm, a new encoding and decoding method was constructed to solve the upper model. The lower level model was solved based on the Frank-Wolfe algorithm. Based on the data of a regional expressway network in Guizhou Province, the above models and algorithms were verified and analyzed. The results show that under certain resource and budget constraints, considering the difference in road section recovery degree can improve the road network performance resilience by 32.62%. Considering the difference in road section recovery speed can improve the road network performance resilience by 10.17%. The sensitivity analysis shows that taking into consideration the difference in road section recovery speed can improve the marginal benefits of increasing the number of recovery resources for improving road network performance resilience, recovery speed resilience, and weighted resilience by 12.69%, 5.47%, and 22.93% respectively. Considering the difference in road section recovery degree helps balance the improvement of road network performance resilience and the reduction of recovery speed resilience caused by the increase of recovery budget, so as to ensure the road network recovery performance. Therefore, it is important to consider the recovery differences in different road sections for road network recovery decision.

Keywords

• road traffic network
• recovery decision optimization
• bi-level programming model
• network resilience
• genetic algorithm