Significant advancements have been made in autonomous driving technology, particularly in multi-lane autonomous driving platoons, which have become a key research topic aimed at optimizing space-time road resources utilization and alleviating traffic congestion. With the application of rational lane allocation and effective platoons strategies, vehicle waiting times and queue lengths can be substantially reduced. The objective of this paper is to design a space-time resources allocation method for multi-lane autonomous driving platoons. An integrated approach is proposed to manage Connected and Automated Vehicles (CAVs) platoons, combining lane allocation based on signal duration with model-predictive control. By accommodating dynamic fluctuations in traffic demands, the methodology enables the passive adaptation of limited road resources, providing both theoretical and practical guidance toward enhancing intersection capacity and efficiency. The proposed method was applied to the intersection of Yizhuang Kechuang Road and Jinghai Road in Beijing, and simulations were conducted by SUMO software. Results on core quality metrics such as average vehicle delay, average parking duration, and average queue length were significantly improved after algorithm optimization.