Support Vector Regression Ship Motion Identification Modeling Based on Grey Wolf Optimizer

In order to improve the accuracy of ship motion identification modeling, the grey wolf optimizer (GWO) algorithm is used to optimize the hyperparameters of support vector regression(SVR). GWO-SVR identification algorithm is proposed and applied to black box identification modeling of ship motion. All the 15° / 15° zigzag test and part of 5°, 20° turning tests data are used as training data to train hyperparameters of SVR. The trained prediction model is applied to predict the whole 20° turning test and 20° / 20° zigzag test data. Compared with the prediction results of SVR based on particle swarm optimization (PSO-SVR), the prediction results of GWO-SVR algorithm show that the proposed algorithm has the advantages of smaller prediction errors, faster convergence speed and better generalization performance.