In order to realize the collaborative optimization of braking energy recovery efficiency and driving stability of distributed drive electric vehicles, a composite braking hierarchical control strategy integrating an anti-lock braking system (ABS) coordination mechanism was proposed. The strategy adopted the upper and lower double-layer optimization architecture. The upper layer utilized a variable-ratio braking force distribution strategy to make the actual distribution curve approximate the ideal braking force distribution curve, thereby maximizing tire-road adhesion utilization. The lower layer consisted of a dynamic coordination control mechanism for electric and hydraulic braking based on slip rate, achieving precise coordination between ABS system and regenerative braking while ensuring braking stability. Then the particle swarm optimization algorithm was introduced to optimize the membership function of fuzzy controller offline, and the regenerative braking force distribution coefficient was dynamically optimized with battery state of charge (SOC), braking strength and vehicle speed as input variables, so as to further improve the energy recovery efficiency under complex working conditions. Based on Matlab/Simulink simulation verification, the optimized SOC contribution rates under NEDC and FTP75 conditions were 17.81% and 18.53% respectively. The energy recovery rate of low adhesion road surface was 52%, and that of μ-split surface was 23.9%. The analysis results show that the strategy can effectively improve the braking stability and energy recovery rate.