<p>With the popularization of information technology and the acceleration of modern life, the online ride-hailing industry has been flourishing in China. This paper first explores the traditional order dispatch mode, identifies its contradictions, and analyzes their underlying causes. Next, a novel dispatch mode is proposed based on a batch-delay matching model. The PLTS-Hungarian algorithm is employed to match orders clustered by a clustering algorithm. Subsequently, evaluation criteria for dispatch modes are established. Finally, order allocation results are obtained using both the traditional and proposed modes. Comparative analysis demonstrates the validity and practicality of the proposed mode. Compared with the objective function values before and after optimization, the new mode improves overall platform efficiency, reflected by increased average pick-up time and higher average order efficiency coefficient. Meanwhile, computational complexity is significantly reduced. In addition, the applicability of the proposed model is verified through sensitivity analysis, and its performance under both high and low demand scenarios is examined. This approach provides valuable guidance for similar scheduling and matching problems.</p>