Mechanism data coupling driven variable working condition zero sample mechanical fault diagnosis method

The invention relates to a mechanical parameter mode identification method, in particular to a mechanism data coupling driven variable working condition zero sample mechanical fault diagnosis method, which is used for solving the problem that although a domain generalization network can realize new working condition data-free fault diagnosis, the model diagnosis performance depends on the distribution of source domain data, and the fault diagnosis efficiency is low. And working condition information and data distribution conditions of new working conditions are not considered, so that the generalization performance is insufficient, and the diagnostic capability is poor. According to the mechanism data coupling driven variable working condition zero sample mechanical fault diagnosis method, when a working condition conversion relation is established, a fault sequence conversion model based on a quadratic polynomial function is proposed, real data of an old working condition can be effectively converted into ne