基于遗传优化的水面无人船艇神经网络智能控制

利用GA智能优化算法和RBF神经网络逼近算法设计了一种USV运动滑模理想跟踪控制方法.首先利用改进的遗传算法对RBF网络参数进行在线寻优以进而提高其逼近性能.其次,将学习速度较快的局部RBF神经网络对滑模控制设计中存在的船舶运动系统函数不确定项进行逼近,使得由于滑模面的不间断切换引起的控制输入抖振问题得到有效地解决.对比实验说明了在同等条件下,上述智能控制系统稳定时间更快,超调量更小,以及输入舵角更平滑.

Genetic Optimized RBF Neural Network Based Intelligent Control Algorithm of USV

The intelligent tracking algorithm of Unmanned Surface Vehicle was proposed based on the optimized RBF network.By modifying the adaptation value and the mutation probability,it is possible to solve the problem of premature convergence of the GA which was used to optimize the network parameters online to solve the problem,and then improve its approximation performance.And then,the RBF neural network,with the advantage of learning fast,was used to approximate the uncertainties of the function of the USV motion system during the ideal Sliding Mode Control designing so that the control input chattering problem caused by the uninterrupted switching of the sliding surface is effectively overcome.A comparative study shows that under the same conditions,the stabilization time of the intelligent control system is faster,the average overshoot is smaller,and the input rudder angle is smoother.