To address the scheduling problem in flexible assembly job shops where sublot processing, transportation, and assembly kitting are highly coupled, existing studies rarely consider the coordinated optimization of sublot flow, transportation resource allocation, and assembly kitting constraints. Therefore, this paper investigates the Flexible Assembly Job Shop Batch Flow Scheduling Problem with Transportation Constraints (FAJSP-LS-T). A comprehensive scheduling model integrating finite transportation resource constraints and batch flow scheduling is developed. Furthermore, a hybrid intelligent optimization algorithm combining Q-learning and NSGA-II is proposed, which dynamically guides the directed selection of crossover and mutation operators and incorporates a local search operator to enhance the local exploitation capability around the Pareto front. Experimental results demonstrate that the proposed hybrid algorithm significantly outperforms traditional multi-objective evolutionary algorithms in minimizing both makespan and total energy consumption. The study provides strong theoretical support for achieving efficient integrated production–logistics scheduling in flexible assembly job shop environments.