Intelligent interactive system has been widely used in urban rescue and daily life. The work aims to explore the inherent mechanism and law of collaborative efficiency improvement in human-autonomy teaming, so as to construct a compatibility theoretical model of human-autonomy teaming and apply it to the design and optimization of human-autonomy teaming in complex systems. Interdisciplinary cross-analysis and literature research were adopted to search and analyze the literature in different fields, such as cognitive psychology, engineering psychology, and computer science. The factors affecting the human-autonomy collaborative efficiency were sorted out systematically from the level of interface design and automation, operator status, ability of situation awareness, human-autonomy relationship and interactive mode. Through the shared mental model, naturalistic decision model and multidimensional representation model, the mechanism for human-autonomy collaboration was discussed and analyzed. The problems and shortcomings of the current research were pointed out. On this basis, a systematic theoretical model of human-autonomy teaming compatibility in intelligent interactive system was proposed. The model analyzed the characteristics of humans, agents, environment, and tasks and discussed the inherent rules and mechanisms of human-autonomy teaming from the perspective of physical compatibility, cognitive compatibility, and emotional compatibility, thus effectively improving the system efficiency and user experience of human-autonomy teaming and avoiding the occurrence of safety accidents. The compatibility model of human-autonomy teaming in intelligent interactive systems can be used in the theoretical and applied research of human-autonomy teaming collaborative efficiency improvement in the future and can also be directly used in the design and optimization of intelligent human-autonomy interactive systems, especially the design and optimization of human-autonomy teaming in the era of intelligence. It can meet the needs of the efficient collaborative design of complex human-autonomy systems (e.g., the cockpits of commercial airplanes and the central control rooms of nuclear power plants) and provide a scientific basis for the establishment of natural and efficient human-autonomy collaborative operation interaction.