Accurate traffic flow prediction is essential for highway operations, especially during holidays when surging traffic poses significant challenges. This study focuses on holiday traffic and introduces a spatiotemporal cross-attention network (ST-Cross-Attn) that combines a bidirectional convolutional LSTM (Bi-ConvLSTM) with a cross-attention module to jointly predict toll station inbound flow and outbound flow. Under the multi-task learning framework, the model shares spatial–temporal features between inbound flow and outbound flow, enhancing their representations and improving multi-step prediction accuracy. Using three years of highway traffic flow data during Labor Day from Shandong, China, ST-Cross-Attn outperformed eight state-of-the-art benchmarks, achieving an average improvement of 4.34% in inbound flow prediction and 2.3% in outbound flow prediction. Extensive ablation studies further confirmed the effectiveness of the model’s components and multi-task learning framework, demonstrating its potential for reliable holiday traffic forecasting.

To validate the validity of a driving simulation platform, this study constructs a dual dataset using field data and driving simulation experiments on an urban arterial road. Comprehensive validation is conducted from both subjective perception and objective quantification perspectives. Subjectively, questionnaires assessed simulation fidelity, driving workload, and physiological symptoms. Objectively, five core car-following indicators (speed, acceleration, relative speed, headway, and time headway) were analyzed using dynamic time warping and statistical methods. The results demonstrate that the simulated scenarios exhibit high subjective fidelity, reasonable task loads, and controllable motion sickness risks. Objectively, dynamic time warping confirms strong temporal pattern similarity, with waveform consistency proportions across core indicators ranging from 91.67% to 100.0%. Macroscopically, satisfactory relative aggregate similarity is demonstrated, with relative difference between means consistently constrained within a 20% threshold. However, strict absolute behavioral validity is unsupported due to significant statistical differences in sequence means, establishing clear boundary constraints for trend replication. It is speculated that systematic biases are primarily related to differences in risk perception within virtual environments.