To tackle the problem of whether the placement of variable message signs (VMS) on a highway section requires optimization, an evaluation model was put forward based on a dynamic fuzzy neural network. Firstly, taking into account the complex and dynamic traffic system of highways, a corresponding evaluation indicator system was established. Secondly, to evaluate the impact of VMS placement on traffic conditions in a more realistic manner, the differences among various management strategies and the non-management strategy were utilized as combined feature data, and the evaluation model was applied to conduct evaluations at the lane level. Finally, a traffic congestion scenario in the highway merge area was constructed, and the effectiveness of the evaluation model was verified through traffic simulation. The results demonstrate that the proposed evaluation model can assess the placement of VMS from multiple traffic indicator perspectives, providing a more comprehensive evaluation that conforms to the trends of individual indicators.