Ship roll motion prediction based on ℓ1 regularized extreme learning machine.

Area of Science:

• Naval Architecture and Marine Engineering
• Computational Intelligence
• Time Series Analysis

Background:

• Accurate prediction of ship roll motion is crucial for maritime safety and operational efficiency.
• Traditional methods often struggle with complex dynamics and real-time prediction accuracy.
• Overfitting and underfitting can significantly degrade the performance of predictive models.

Purpose of the Study:

• To propose a novel Extreme Learning Machine (ELM) based method for enhanced ship roll motion prediction.
• To improve prediction accuracy and model robustness by optimizing ELM structure.
• To validate the proposed method using real-world ship roll motion data.

Main Methods:

• Input selection from roll motion time series using the Lipschitz quotient method.
• Determination of hidden layer nodes via an ℓ1 regularization technique.
• Solving the ℓ1 regularized ELM using the least angle regression (LAR) algorithm.

Main Results:

• The proposed method demonstrates improved accuracy in ship roll motion prediction.
• Optimization techniques effectively prevent over or underfitting, leading to robust models.
• Experimental validation confirms the method's effectiveness on real measured data.

Conclusions:

• The developed ℓ1 regularized ELM offers a superior approach for ship roll motion prediction.
• The combination of Lipschitz quotient and ℓ1 regularization enhances model performance.
• This method provides a valuable tool for real-time maritime safety and operational planning.