Intelligent Sea States Identification Based on Maximum Likelihood Evidential Reasoning Rule

Area of Science:

• Marine Engineering
• Naval Architecture
• Artificial Intelligence

Background:

• Ship navigation requires frequent adjustments to propulsion control strategies based on changing sea states for safety and stability.
• Timely and effective sea state identification is critical for ensuring overall ship safety during navigation.

Purpose of the Study:

• To develop an intelligent sea state identification model.
• To utilize propeller ventilation type as a key indicator for sea state classification.
• To enhance the accuracy and reliability of sea state identification for maritime applications.

Main Methods:

• Construction of a data-driven maximum likelihood evidential reasoning (MAKER) model, accounting for input feature interdependencies.
• Optimization of MAKER model parameters using a genetic algorithm (GA) to improve evaluation accuracy.
• Training and validation of the MAKER model using simulated experimental data.

Main Results:

• The developed MAKER-based model accurately identifies propeller ventilation types.
• The model demonstrates enhanced accuracy in intelligent sea state identification compared to traditional methods.
• The study validates the effectiveness of the proposed intelligent sea state identification approach.

Conclusions:

• The MAKER rule-based intelligent model offers a more accurate method for identifying propeller ventilation types.
• This approach enables effective and intelligent identification of sea states, contributing to maritime safety.
• The integration of data-driven modeling and optimization techniques proves beneficial for complex marine systems.