In this paper, we address the low matching efficiency and information asymmetry in vehicle and cargo matching by constructing a model based on actual complex business scenarios. This model accurately matches vehicle and cargo characteristics, improves overall matching efficiency, reduces idling rates, and optimizes matching revenue and transportation costs for drivers. An improved ant colony algorithm is proposed, incorporating a dynamic pheromone updating strategy, adaptive selection probability adjustment, and path diversity evaluation to enhance global search efficiency. Applied to real data from the Full Truck Alliance platform, the experimental results show the improved algorithm outperforms traditional and other intelligent optimization algorithms in efficiency and stability. This work offers a new solution for vehicle and cargo allocation and a valuable reference for applying intelligent optimization algorithms in logistics.