Time series forecasting serves a critical function in domains such as energy, meteorol-ogy, and power systems by leveraging historical data to predict future trends. Howev-er, existing methods often prioritize long-term dependencies while neglecting the inte-gration of local features and global patterns, resulting in limited accuracy for short-term predictions of non-stationary multivariate sequences. To address these challenges, this paper proposes a time series forecasting model named VBTCKN based on Variational Mode Decomposition and a Dual-channel Cross-attention Network.