论文部分内容阅读
自治式水下机器人执行使命时,沿规划路径运动的导航,特别是围绕被搜索到的目标物作圆周运动的导航,是一种强非线性的控制问题、传统方法解决这一问题很困难,神经网络为非线性控制问题的实施提供了快速、准确的工具。利用神经网络构造的非线性控制器为自治式水下机器人的导航和定位提供了一种新的途径,网络经训练后能适应混有噪声的非线性环境。本文综合利用前惯和反馈控制器的优点,把GESA算法应用于自动增强的函数链网络,首先对自治式水下机器人实施前馈控制,再利用神经网络构成的PID控制器来提高控制系统的控制精度,使自治式水下机器人沿规划轨线运动。
Autonomous Underwater Vehicle (ROBO) performs a mission, navigation along a planned path, and especially navigation around circular motion of the object being searched, is a strongly non-linear control problem that is difficult to overcome by conventional methods, Neural networks provide a fast and accurate tool for the implementation of nonlinear control problems. The nonlinear controller constructed by neural network provides a new way for the navigation and localization of autonomous underwater vehicles. The network can be trained to adapt to the non-linear environment with noises. In this paper, the advantages of the former inertial feedback controller are comprehensively used. The GESA algorithm is applied to the automatically enhanced function chain network. Firstly, the autonomous underwater robot is fed with feedforward control, then the neural network PID controller is used to improve the control system Control accuracy allows autonomous autonomous underwater robots to move along planned trajectories.