Recent foundation image and video generation models offer strong generalization and controllability, but their direct application to embodied scenarios is limited by requirements for multi-view consistency, geometric coherence, and robot embodiment constraints. Existing methods t…
To tackle price volatility in distribution network material procurement and improve precision in setting reference prices, this study introduces two hybrid machine learning methods for price forecasting. Materials are first classified using the ABC method. CNN-SVM and TCN-GRU hyb…