Compromised Decision-Making Approach to Speed Limit Value of Hard Shoulder Running at Freeway Merge Areas Using Integrated Entropy Weight Method-VIKOR Method

The hard shoulder running strategy is a valuable measure for alleviating traffic congestion. This study aims to propose a speed limit decision-making approach for the hard shoulder running strategy in freeway merging areas. Traffic simulation experiments are conducted using the SUMO simulator to analyze the impact of the hard shoulder running strategy on road capacity. The study introduces a transition section between the merging segment and the hard shoulder operation segment to prevent traffic conflicts and assesses the appropriate length of the transition section in the merging area. Ultimately, the entropy weight method (EWM)-VIKOR method is used to determine the optimal speed limit thresholds for hard shoulder operation in the merging area. Simulations are carried out using the Jingxiong freeway in China as a case study. The results show that the hard shoulder running strategy can increase the capacity of the Jingxiong freeway’s four-lane section by 22.5%, with the optimal traffic volume range for implementing this strategy being 8,000–9,800 passenger cars per hour (pcph). In the merging area, it is recommended that the transition section be longer than 5 km. Taking a traffic volume of 8,800 pcph as an example, the optimal speed limits for the five lanes from the innermost lane to the hard shoulder lane are set at 110 km/h, 110 km/h, 100 km/h, 100 km/h, and 90 km/h, respectively. This method can be applied to freeways with different lane counts and speed limits, providing theoretical support for setting reward values in dynamic speed control on freeways based on deep reinforcement learning.