In order to create a simulation test platform to effectively test the key technologies of intelligent ships such as guidance, navigation and control technology, this study uses system identification technology to identify the parameters of the Nomoto motion model of an intelligent ship with high precision.

A hybrid parameter identification method is proposed by fully combining the advantages of the extended state observer (ESO) and the robust weighted least square support vector regression algorithm (RW-LSSVR), our previously well-evaluated identification method. The ESO-based state estimator is applied to calculate immeasurable states using measurable states and the second-order linear Nomoto model. To evaluate the proposed approach, models of two vessels with predefined parameter values are employed for simulation tests.

The proposed approach not only estimates immeasurable states with high accuracy, but also ensures good performance in steering model parameter identification, with values very close to the nominal values.

The proposed ESO-based identification method shows good generalizability and can effectively provide satisfactory estimates of immeasurable states, making it highly applicable to parameter identification.