In recent years, rapid development has been continuously observed in China's manufacturing industry, where AGVs automated guided vehicles have been increasingly adopted by enterprises as core equipment in intelligent logistics systems. To ensure the efficiency of warehouse operations, addressing the issue of transportation path conflicts among AGVs has garnered growing attention from researchers. A literature review on the issue of multi-AGV path conflicts in warehouses was conducted from two perspectives. First, from the perspective of conflict types, the research problems were categorized into collision problems and deadlock problems, and the current state of research on multi-AGV collision avoidance strategies under different conflict types was analyzed. Second, from the perspective of model-solving algorithms, the study divides the approaches into heuristic algorithms and reinforcement learning algorithms, analyzing their application in multi-AGV path conflict issues in warehouses in recent years. Finally, the existing literature was summarized, and future directions for addressing multi-AGV path conflicts in warehouses were proposed.