In recent years, new models of testing machines have emerged in the aviation industry. In order to maintain model competitiveness, it is necessary to improve flight efficiency while ensuring safety, accelerate the progress of new model airworthiness certification flight testing, and shorten the development cycle. During the flight test process, subjects such as stall test, minimum ground clearance speed test, landing gear shimmy test, flutter test, natural icing test, high temperature, plateau, and high cold test are difficult and risky, and the safety and efficiency of the test must be well guaranteed, posing a huge challenge to monitoring capabilities. In order to improve flight test efficiency and personnel utilization while ensuring flight test safety, this article proposes a multi-machine and multi-subject safety monitoring technology based on expert libraries, which breaks through key technologies such as building a safety monitoring system architecture based on rapid response engineering, designing a flight test safety monitoring expert knowledge base, parallel real-time processing of multiple data streams based on streaming computing frameworks, and multi-machine and multi subject safety monitoring based on task themes, Develop a multi-aircraft and multi subject safety monitoring system based on the expert knowledge base of flight test safety monitoring, and build an expert knowledge base that integrates analysis algorithms, alarm diagnosis, decision support, aircraft information and other flight test safety monitoring knowledge. This achieves loose coupling of the safety monitoring system's business logic data visualization display, and can quickly respond to changes in demand. At the same time, it supports real-time switching of various monitoring methods such as "single terminal, multiple machines, and multiple subjects", Providing a new way to optimize the allocation of human resources for safety monitoring.