1. Research on high precision online prediction of motion responses of a floating platform based on multi-mode fusion.
- Author
-
Wang, Jianwei, Jin, Xiaofan, He, Ze, Wang, Yuqing, Liu, Xuchu, Chai, Jiachen, and Guo, Rui
- Subjects
- *
PREDICTION models , *FORECASTING , *OCEAN , *SIGNALS & signaling , *NOISE - Abstract
• A multi-mode fusion method is proposed to predict the motion of platform. • The signals are decomposed into approximately stationary signals by VMD method. • The proposed method achieves high precision real-time prediction. To address poor prediction accuracy caused by signal fluctuation, randomness, and noise interference in motion response prediction for floating platforms, a method combining Variational Modal Decomposition (VMD) and multi-model fusion is proposed. Initially, raw data undergoes VMD preprocessing, yielding Intrinsic Modal Functions (IMFs). These IMFs, along with ocean environmental data, are utilized in Long Short-Term Memory (LSTM), Bi-Directional Long Short-Term Memory (BiLSTM), and Gated Recurrent Units (GRU) models. Subsequently, Bayesian Linear Regression (BLR) is applied to the combined prediction results and original features for final prediction. Experimental results demonstrate superior stability and accuracy compared to other model prediction methods, validating the proposed approach's efficacy and feasibility in tackling prediction challenges. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF