A conflict-aware extension of the cell transmission model for consistent intersection delay simulation

Macroscopic traffic flow models such as the Cell Transmission Model (CTM) are widely used in large-scale traffic simulation; however, discrepancies between macroscopic and microscopic representations persist at unsignalized intersections with conflicting flows. While microscopic simulations predict conflict-induced delays even under low-demand conditions, standard CTM formulations exhibit a systematic macroscopic simulation artifact at unsignalized intersections: as long as flow-dependent capacity constraints are inactive, disturbed movements traverse the intersection at free-flow speed, leading to unrealistically delay-free behavior. This paper addresses this inconsistency by proposing a conflict-aware extension of the CTM (CA-CTM) based on a minimal structural modification of the sending function. Rather than representing conflicts solely through flow-dependent maximal outflow constraints, the effective free-flow speed in selected cells is reduced as a function of opposing demand. This preserves the core structure of the CTM while increasing computational effort only by a constant factor, thereby improving consistency with microsimulation dynamics across different demand regimes. Using controlled simulation-based case studies in an isolated intersection setting and a stylized urban network, we show that standard CTM formulations can be calibrated to align reasonably well with microscopic reference results within individual scenarios but fail to generalize across varying conflict levels. In contrast, the proposed CA-CTM remains sensitive to conflict effects across demand regimes, suggesting that the discrepancies between standard CTM formulations and microscopic reference models under conflicting flows are structural rather than parametric and can be substantially reduced by a minimal intervention.