Journal of Shanghai Maritime University, Volume. 46, Issue 2, 23(2025)

Fuel consumption prediction model for sail-assisted ships based on hybrid kernel function ARS-SVR

LIU Yize, MA Ranqi, RUAN Zhang, and HUANG Lianzhong*
Author Affiliations
  • School of Turbine Engineering, Dalian Maritime University, Dalian 116026, Liaoning, China
  • show less

    In order to effectively predict the fuel consumption of ships, a ship fuel consumption prediction model based on a hybrid kernel function is proposed. The support vector regression (SVR) models of the radial basis function (RBF) and the polynomial single-kernel function are constructed respectively, and adaptive random search (ARS) algorithm is used to optimize them. On this basis, the ship fuel consumption prediction model based on the hybrid kernel function ARS-SVR is established. A sail-assisted very large crude carrier (VLCC) is used as the research object, and the ship fuel consumption prediction is carried out based on the real ship monitoring data. The results show that compared with the single RBF and the polynomial single-kernel ARS-SVR, the root mean square error of the prediction result by the model based on the hybrid kernel function ARS-SVR is reduced by 19.8% and 30.2%,respectively. The proposed ship fuel consumption prediction model can improve the accuracy of the calculation of the sail-assisted ship fuel consumption,and is helpful to optimize the energy efficiency of ships and improve the management technology.

    Tools

    Get Citation

    Copy Citation Text

    LIU Yize, MA Ranqi, RUAN Zhang, HUANG Lianzhong. Fuel consumption prediction model for sail-assisted ships based on hybrid kernel function ARS-SVR[J]. Journal of Shanghai Maritime University, 2025, 46(2): 23

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Received: Apr. 26, 2024

    Accepted: Aug. 22, 2025

    Published Online: Aug. 22, 2025

    The Author Email: HUANG Lianzhong (lzhuang@dlmu.edu.cn)

    DOI:10.13340/j.jsmu.202404260076

    Topics