面向实船运动特性的轨迹发生器设计

Trajectory generator design for real ship motion characteristics

为提高轨迹发生器生成的仿真数据与实船运动状态的符合性,设计了一种面向实船运动特性的船舶轨迹发生器。利用GPR-DBSCAN算法,将实测数据识别分割成不同运动状态,再结合船舶运动学和动力学方程,建立了不同运动状态的船舶运动数学模型,并给出了姿态速度和位置(AVP)、惯性测量单元(IMU)数据生成算法。最后利用非线性回归辨识模型中的未知参数,完成船舶轨迹发生器的构建。仿真和实测数据验证结果表明,构建的船舶轨迹发生器与PSINS工具箱中的轨迹发生器相比,位置、航速和航向准确度分别提高了46.73%、6.91%和40.71%,更符合实船运动特性。

To improve the consistency between the simulation data generated by the trajectory generator and the actual ship motion state,a ship trajectory generator is designed for the real ship motion characteristics..Using the GPR-DBSCAN algorithm,the measured data is identified and divided into different motion states.Then,combined with ship kinematics and dynamics equations,mathematical models of ship motion for different motion states are established,and attitude velocity and position(AVP)and inertial measurement unit(IMU)data generation algorithms are provided.Finally,nonlinear regression is used to identify unknown parameters in the model and complete the construction of the ship trajectory generator.Simulation and experimental data verification show that the constructed ship trajectory generator has improved position,speed,and heading accuracy by 46.73%,6.91%,and 40.71%,respectively,compared to the trajectory generator in the PSINS toolbox,which is more suitable for the motion characteristics of real ships.