高速船运动在线联机预报及其控制仿真

Online Prediction and its Control Simulation of High-speed Craft Motion

船舶在航行时由于航行工况和环境变化多端 ,受力十分复杂 ,使得船舶推进系统具有非线性、时变性和干扰的复杂性等特点。该文基于MATLAB/SIMULINK对某一高速船推进系统进行建模。在此模型中 ,利用神经网络对航行海况进行在线联机预报 ,预报结果传给局域网中另一计算机的模糊控制器 ,对船桨子系统进行在线控制 ,从而改变船舶航速。同时船舶的实时航速被反馈到神经网络预报器 ,进行下一轮预报及其控制 ,如此反复 ,最终实现智能控制。对翼滑艇的仿真研究结果表明 ,应用此种联机控制 ,在各种海况下 ,均能达到良好的控制效果。有理由相信神经网络结合模糊控制在船舶工程中有很好的应用前景。

There are some characteristics such as nonlinear, big inertia and complicated disturbances during navigation due to the changing working situation and environment. A model of high-speed craft propulsion system based on MATLAB/SIMULINK is presented in this paper. In this model, online prediction of the environmental parameters is performed by neural network predictor. Once the prediction is achieved, the parameters then are sent to the Fuzzy Logic Controller, which located in another computer in the same LAN. Meanwhile, the real-time speed is returned to neural network in order to prepare for the next step prediction. With the repeat of the forementioned steps an automatic ship control is realized. The simulation results of a gliding hydro craft under different disturbances show that a powerful capacity of the online control system is achieved whether in calm sea or in rough sea. It's safe to believe that neural network and fuzzy logic have good application foreground in shipbuilding.