A method for modeling ship fuel consumption based on DBO optimization

In order to solve the problems of low prediction accuracy and slow parameter tuning of ship fuel consumption model constructed based on Bidirectional Long Short-Term Memory (BiLSTM), In this paper, a method based on the Dung beetle optimizer (DBO) optimization BiLSTM is proposed to construct a ship fuel consumption model, and the attention mechanism is combined with weighted aggregation features. Firstly, the construction of ship fuel consumption model and parameter optimization method are introduced. Secondly, aiming at the shortcomings of BiLSTM prediction model, the fuel consumption prediction model (DBA) was constructed based on DBO method to optimize parameters. Finally, taking the sailing data of a ship in the ocean as a sample, the above method is used to forecast the ship's fuel consumption, and the results are compared with the BiLSTM model. The experimental results show that the determination coefficient R^2 of the optimized model is 0.0028 higher than that of BiLSTM, and MAE and RMSE are 0.062 and 0.042 lower, respectively, which can provide a certain reference for ship energy conservation and emission reduction. provide a feasible reference scheme for engineering applications.