Search Results

You are looking at 1 - 2 of 2 items for

  • Author or Editor: Hao Dai x
  • All content x
Clear All Modify Search
Xining Zhang and Hao Dai


In recent years, deep learning technology has been gradually used for time series data prediction in various fields. In this paper, the restricted Boltzmann machine (RBM) in the classical deep belief network (DBN) is substituted with the conditional restricted Boltzmann machine (CRBM) containing temporal information, and the CRBM-DBN model is constructed. Key model parameters, which are determined by the particle swarm optimization (PSO) algorithm, are used to predict the significant wave height. Observed data in 2016, which are from nearshore and offshore buoys (i.e., 42020 and 42001) belonging to the National Data Buoy Center (NDBC), are taken to train the model, and the corresponding data in 2017 are used for testing with lead times of 1–24 h. In addition, we trained the data of 42040 in 2003 and tested the data in 2004 in order to investigate the prediction ability of the CRBM-DBN model for the extreme event. The prediction ability of the model is evaluated by the Nash–Sutcliffe coefficient of efficiency (CE) and root-mean-square error (RMSE). Experiments demonstrate that for the short-term (≤9 h) prediction, the RMSE and CE for the significant wave height prediction are <10 cm and >0.98, respectively. Moreover, the relative error of the short-term prediction for the maximum wave height is less than 26%. The excellent short-term and extreme events forecasting ability of the CRBM-DBN model is vital to ocean engineering applications, especially for designs of ocean structures and vessels.

Full access
Guomei Wei, Zhigang He, Yanshuang Xie, Shaoping Shang, Hao Dai, Jingyu Wu, Ke Liu, Rui Lin, Yan Wan, Hang Lin, Jinrui Chen, and Yan Li


Two Ocean State Monitoring and Analyzing Radar (OSMAR071) (7.8 MHz) high-frequency (HF) radars and four moored ADCPs were operated concurrently in the southwestern Taiwan Strait during January–March 2013. Qualitative and quantitative comparisons of surface currents were conducted between the HF radars and the ADCPs. Except for a location probably affected by shallow water and sand waves on the Taiwan Banks, the HF-radar-derived radial currents (radials) showed good agreement with the ADCP measured results (correlation coefficient: 0.89–0.98; rms difference: 0.07–0.13 m s−1). To provide further insight into the geophysical processes involved, the performance of the HF-radar-derived radials was further evaluated under different sea states (sea states: 2–6). It was found that both the data returns of the radar-derived radials and the differences between the radar-derived radials and the ADCP-derived radials varied with sea state. The HF radar performed best at sea state 4 in terms of data returns. The spatial coverage increased rapidly as the waves increased from sea state 2 to 4. However, it decreased slowly from sea state 4 to 6. Second, the radial differences were relatively high under lower sea states (2 and 3) at the location where the best agreement was obtained between the radar and ADCP radials, whereas the differences increased as the sea states increased at the other three locations. The differences between the radials measured by the HF radars and the ADCPs could be attributed to wave-induced Stokes drift and spatial sampling differences.

Restricted access