Performance Evaluation of Adaptive Neuro Fuzzy Inference System-Based Ship Energy Management Strategy Under Real Operating Conditions

Abstract

Traditional fuzzy logic control strategies have a certain subjectivity and lack adaptive capability. In this study, an energy management strategy (EMS) based on adaptive neuro fuzzy inference system (ANFIS) is proposed for a 7500-ton diesel-electric hybrid power bulk carrier. First, a hybrid power system model is designed using MATLAB/Simulink. Then, the fuzzy controller and ANFIS controller are designed under pre-set operating conditions, and the real operating conditions of the target ship are constructed based on real sailing data. Furthermore, a two-layer power allocation for the target ship is conducted. Based on this, EMS based on two-layer fuzzy logic control and two-layer ANFIS are developed. Finally, comparative experiments are carried out under real operating conditions. The results show that in real operating conditions, the adaptability of EMS based on ANFIS is more better; fuel consumption decreases by 4.3%, carbon dioxide (CO2) emissions reduce by 4.5%; battery state of charge (SOC) fluctuation range narrows by 3.68%, initial/final offset decreases by 54.3%; response speed of ANFIS controller is faster than fuzzy controller by 14.83% and 14.92% in first and second layer power allocation respectively.This indicates that EMS based on ANFIS can reduce energy consumption, extend battery life, effectively adapt to different sailing conditions and improve ship navigation efficiency.

Keywords: Adaptive neuro fuzzy inference system, Energy management strategy, Hybrid power ship, Testing and verification, Real operating condition