An EMD-LSTM-SVR model for the short-term roll and sway predictions of semi-submersible

Ye, Yutu; Wang, Lei; Wang, Yiting; Qin, Licheng

doi:10.1016/j.oceaneng.2022.111460

Cited by 29 publications

(8 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The low prediction accuracy of surge motions employing DeepONet and WNO may be attributed to its broad-band spectrum feature. This conclusion is consistent with that given by Ye et al [6], which states that the spectrum bandwidth deteriorates the prediction accuracy of DNNs. To that end, we increased the number of samples generated for all combinations of H s and T z for surge motion in cases 4 and 5.…”

Section: Numerical Resultssupporting

confidence: 93%

“…Data-driven machine learning algorithms like dynamic mode decomposition [2], sparse identification for nonlinear dynamics (SINDy) [3], and deep neural networks (DNNs) [4,5] have emerged as viable and critically enabling methodologies to address the challenges of accurate predictions of complex responses for offshore floating structures [5][6][7][8], and have led to new developments in neural simulators [9][10][11][12][13][14][15][16]. Real-time response prediction usually refers to forecasting future responses within tens of seconds in real-time based on past excitation or responses.…”

Section: Introductionmentioning

confidence: 99%

“…For example, Guo et al [5] predicted the heave and surge of a semi-submersible platform for 46.5s into the future by combining the LSTM cells and several fully connected layers. Furthermore, Ye et al [6] proposed a hybrid empirical mode decomposition (EMD) model for the short-term roll and sway predictions of a semi-submersible platform. The EMD technique decomposed motion time series into a certain number of intrinsic mode functions (IMFs), and each IMF was predicted by a support vector regression model except the first IMF, which was predicted employing LSTM.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Deep neural operators can predict the real-time response of floating offshore structures under irregular waves

Cao¹,

Goswami²,

Karniadakis³

et al. 2023

Preprint

View full text Add to dashboard Cite

The use of neural operators in a digital twin model of an offshore floating structure can provide a paradigm shift in structural response prediction and health monitoring, providing valuable information for real-time control. In this work, the performance of three neural operators is evaluated, namely, deep operator network (DeepONet), Fourier neural operator (FNO), and Wavelet neural operator (WNO). We investigate the effectiveness of the operators to accurately capture the responses of a floating structure under six different sea state codes (3 − 8) based on the wave characteristics described by the World Meteorological Organization (WMO). The results demonstrate that these high-precision neural operators can deliver structural responses more efficiently, up to two orders of magnitude faster than a dynamic analysis using conventional numerical solvers. Additionally, compared to gated recurrent units (GRUs), a commonly used recurrent neural network for time-series estimation, neural operators are both more accurate and efficient, especially in situations with limited data availability. To further enhance the accuracy, novel extensions, such as wavelet-DeepONet and self-adaptive WNO, are proposed. Taken together, our study shows that FNO outperforms all other operators for approximating the mapping of one input functional space to the output space as well as for responses that have small bandwidth of the frequency spectrum, whereas for learning the mapping of multiple functions in the input space to the output space as well as for capturing responses within a large frequency spectrum, DeepONet with historical states provides the highest accuracy.

show abstract

Section: Numerical Resultssupporting

confidence: 93%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Deep neural operators can predict the real-time response of floating offshore structures under irregular waves

Cao¹,

Goswami²,

Karniadakis³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Considering that various periodic fluctuation components hide in the original customer electrical load curves, time series decomposition methods such as empirical mode decomposition (EMD) and variational mode decomposition (VMD) can effectively help improve the accuracy of short-term load forecasting (Bedi and Toshniwal, 2018;Ye et al, 2022). However, some intrinsic mode components obtained by decomposition may have similar frequencies and fluctuation patterns.…”

Section: Introductionmentioning

confidence: 99%

CEEMDAN-CatBoost-SATCN-based short-term load forecasting model considering time series decomposition and feature selection

et al. 2023

View full text Add to dashboard Cite

The rapidly increasing randomness and volatility of electrical power loads urge computationally efficient and accurate short-term load forecasting methods for ensuring the operational efficiency and reliability of the power system. Focusing on the non-stationary and non-linear characteristics of load curves that could easily compromise the forecasting accuracy, this paper proposes a complete ensemble empirical mode decomposition with adaptive noise–CatBoost–self-attention mechanism-integrated temporal convolutional network (CEEMDAN-CatBoost-SATCN)-based short-term load forecasting method, integrating time series decomposition and feature selection. CEEMDAN decomposes the original load into some periodically fluctuating components with different frequencies. With their fluctuation patterns being evaluated with permutation entropy, these components with close fluctuation patterns are further merged to improve computational efficiency. Thereafter, a CatBoost-based recursive feature elimination algorithm is applied to obtain the optimal feature subsets to the merged components based on feature importance, which can effectively reduce the dimension of input variables. On this basis, SATCN which consists of a convolutional neural network and self-attention mechanism is proposed. The case study shows that time series decomposition and feature selection have a positive effect on improving forecasting accuracy. Compared with other forecasting methods and evaluated with a mean absolute percentage error and root mean square error, the proposed method outperforms in forecasting accuracy.

show abstract

“…In this category, for instance, [23] EMD has been used with the RBFNN model and improved sway prediction despite requiring a long observation length. The literature in this area is relatively resourceful and promising, such as [24], [25], and [26], where the combination of classic/intelligent regressors comprising AR, SVR, and LSTM operating on the decomposed signal as the hybrid models, have dramatically improved prediction of stationary signals for the longer horizons. Nevertheless, all decomposition techniques and nonlinear/intelligent regressors possess their specific challenges, such as boundary effect and lack of mathematical foundation for EMD, fixed basis function for wavelet, and high convergence toward local minima for support vector regression (SVR), even though they have been recognized as the practical solutions to date [24].…”

Section: Introductionmentioning

confidence: 99%

Online Short-term Ship Response Prediction with Dynamic Buffer Window Using Transient Free Switching Filter

Majidiyan,

Enshaei

et al. 2023

Preprint

View full text Add to dashboard Cite

Ship response prediction is one of the emerging interests in seakeeping, given the extensive range of applications for autonomous control of marine vehicles. In particular, the shortterm prediction and online updates of the ship response have been in the practical attention span. Despite a body of studies on different predictors, the asymptotic properties of estimators concerning the time series sample size have not been addressed specifically. The predictors only have been analysed for a fixed observation window regardless of the intrinsic statistical characteristics obtainable in the time series. To this end, the current research has considered the performance of two nonlinear and linear based regressors: support vector regression (SVR) and the adaptive Seasonal Auto-Regressive and Integrated Moving Average (SARIMA) model on the time series data obtained from a simulated semisubmersible platform in different sea states. The experiment demonstrated that the prediction accuracy highly depends on the observation window length with respect to the statistical properties of the responses in various wave conditions. Therefore, an innovative filter using the weighted average of the proposed functions has been introduced to dynamically adjust the buffer window based on the signal statistical characteristics. The promising results showed that the filter not only could contribute to promoting any predictors used in autonomous onboard decision-making, guidance, and stabilization systems for present and future intelligent onboard control systems, but the proposed mechanism can be employed in dynamic systems in other disciplines too.

show abstract

An EMD-LSTM-SVR model for the short-term roll and sway predictions of semi-submersible

Cited by 29 publications

References 22 publications

Deep neural operators can predict the real-time response of floating offshore structures under irregular waves

Deep neural operators can predict the real-time response of floating offshore structures under irregular waves

CEEMDAN-CatBoost-SATCN-based short-term load forecasting model considering time series decomposition and feature selection

Online Short-term Ship Response Prediction with Dynamic Buffer Window Using Transient Free Switching Filter

Contact Info

Product

Resources

About