Research on Dynamic Adaptability Prediction of Civil Aviation Vocational College Major Group Based on Multimodal Data of Curriculum System and VMD Transformer

The digital and intelligent transformation of the civil aviation industry has driven a drastic change in the talent demand structure, necessitating a workforce capable of navigating smart airports, aviation logistics, maintenance, navigation and communication systems. As the core carrier of civil aviation technical and skilled talent cultivation, the dynamic adaptability of the curriculum system and industry job requirements directly determines the quality of talent cultivation in the civil aviation vocational professional group. The current construction of civil aviation vocational college professional groups faces problems such as lagging curriculum system updates, lack of quantitative basis for adaptability evaluation, and neglect of multi-source data coupling characteristics and dynamic temporal patterns in demand fore casting. Traditional single dimensional data statistics and linear prediction methods are difficult to accurately capture the dynamic changes in industry demand and the multimodal response characteristics of the curriculum system. Variational Mode Decomposition (VMD) can effectively decompose the multi-scale features of nonstationary temporal data, and the Transformer model can deeply explore the spatiotemporal correlations and complex coupling relationships of multimodal data through its self-attention mechanism. By integrating VMD and Transformer models, this article constructs a dynamic adaptability prediction model for civil aviation vocational majors based on multimodal curriculum data, effectively overcoming the limitations of traditional prediction methods in this technically evolving field.