| With the increasing demand for maritime cargo transportation,the impact of ship navigation safety on economic trade,military affairs,and the environment is self-evident.An accurate and reliable mathematical model of ship motion is the key to effectively analyzing ship maneuverability and ensuring intelligent navigation of ships.The ways to establish a mathematical model of ship motion include mechanism modeling and identification modeling.With the development of emerging intelligent identification algorithms and control theory,identification modeling is widely used due to its low cost,high accuracy,and wide applicability.The key to establishing an accurate mathematical model of ship motion is to obtain accurate model parameters.This thesis is based on the Simulated Annealing Particle Swarm Optimization(SAPSO)algorithm and uses the gray box identification method to study the mathematical model of ship motion,in order to achieve better parameter identification and motion prediction results.Particle swarm optimization(PSO)is a popular swarm intelligence algorithm at present.Inspired by birds’ foraging activities,this algorithm regards particles as birds and regards the optimal value as food,it simulates the information sharing among members of a bird flock during the predation process and iterates the algorithm for optimization.This thesis introduces the idea of simulated annealing algorithm on the basis of particle swarm optimization,improves it,and applies it to the problem of parameter identification of ship motion mathematical model.This thesis first establishes a mathematical model of ship motion for the whole body and its simplified model,introduces the basic principles of Least Squares(LS)and PSO algorithm,uses the parameters identified by LS as theoretical reference values,provides a certain search range for particle optimization,and improves the PSO algorithm using simulated annealing algorithm to obtain the SAPSO algorithm,which improves the problem of particle swarm being easily trapped in local optima.Three types of ship models and three different sources of experimental data were selected as research objects and samples for the simplified forms of the established three degree of freedom and four degree of freedom integer models.Firstly,using the characteristics of no measurement noise and environmental interference in the test data collected by the navigation simulator,the zigzag test data of the large tanker 3045 was identified and modeled,and the feasibility and generalization of the SAPSO algorithm were analyzed;Then,identification modeling was conducted using the zigzag test and turning test data of the KCS ship model pool test,which is a standard model for international container ship research.Considering the presence of certain measurement noise interference in the test data,further analysis was conducted on the feasibility and generalization of the SAPSO algorithm;Finally,this thesis takes the real ship YUKUN as the research object,takes its 10 °/10 °,20 °/20 ° zigzag test data in the real ship sea area as the identification samples,and conducts system identification research on the linear and nonlinear ship motion mathematical model.Based on the parameters obtained from the identification model,the 25 ° turning test data is generalized and predicted.The effectiveness and generalization of the algorithm proposed in this thesis are verified through the comparison of the prediction error of the real ship test.Through the research in this thesis,the prediction results obtained by the SAPSO algorithm are superior to those of the PSO algorithm,indicating that the improved SAPSO algorithm has good generalization ability and can be used for grey box identification modeling of ship motion mathematical models.It also provides a reference for determining the hydrodynamic parameters of ship motion mathematical models,thus providing theoretical support for establishing high-precision motion mathematical models of intelligent ship navigation control simulation testing systems. |
