水下机器人动态系统协同建模方法研究

Collaborative Modeling of Underwater Vehicle Dynamic System

以水下机器人动态系统建模问题为背景,提出了基于MATLAB/Simulink与FLUENT协同仿真机制与动态神经网络的水下机器人动态系统协同建模方法。该方法利用在对水下机器人动态行为进行协同仿真过程中在线训练神经网络的方式完成水下机器人流体动力学建模,并以最终训练完成的神经网络及载体6自由度运动方程仿真模块一起构成水下机器人动态系统模型,具有较强工程应用前景。通过与传统建模方法比较,详细论述了协同建模方法的基本原理及其优点,并重点研究了MATLAB/Simulink与FLUENT协同仿真系统的设计方案和实现方法,最后,从水下机器人6自由度运动方程建模、流体动力学建模以及闭环控制系统建模三个方面研究了实现协同建模的具体方法。

Aiming at the dynamic system modeling of underwater vehicle, a new collaborative modeling method was proposed to model the dynamic system of underwater vehicle, which is based on the dynamic neural networks and the Collaborative Simulation mechanism between MATLAB/Simulink and FLUENT, This method completes the hydrodynamic modeling of underwater vehicle by means of online training neural networks during collaboratively simulating dynamics of the underwater vehicle, and constructs the dynamic system model based on the successfully trained neural networks and 6DOF motion simulation module. The fundamental principle and advantages of the collaborative modeling method were discussed through comparison with the traditional modeling method, and the realization scheme of the collaborative simulation system was studied between MATLAB/Simulink and FLUENT. The concrete realization of collaborative modeling method was discussed from the following aspects, including 6DOF motion equation modeling, hydrodynamic modeling and closed-loop control system modeling of underwater vehicle.