Data-driven predictive maintenance and fault diagnosis system

The invention relates to the technical field of airborne electronic equipment inspection and maintenance, in particular to a data-driven predictive maintenance and fault diagnosis system, which comprises a data access and management module used for realizing adaptation and management of data acquisition, effective troubleshooting experience accumulation and management of databases in a historical normal mode and a fault mode, the data access and management module comprises an Internet of Things structure, a fault tree structure and a historical database; the data model module is used for updating a retrieval model based on a natural language processing technology and tracking a model application state, and the data model module comprises a diagnosis knowledge base and an early warning model base; and the data application module is used for a data analysis task, a maintenance/repair scheme and a task statistical report. According to the invention, data-driven fault risk early warning and parameter threshold ra