【摘要】 针对欠驱动自主水下航行器(AUV)在复杂海洋环境中水平面轨迹跟踪控制难度大、抗干扰能力弱的问题,提出一种基于径向基函数(RBF)神经网络的欠驱动AUV反步积分滑模控制(BISMC)轨迹跟踪控制方法。首先,运用反步控制方法设计运动学控制器,得到虚拟控制律和实际控制输入;其次,在动力学控制器中,引入积分滑模控制,考虑系统的不确定因素及可能存在的外部干扰,同时引入RBF神经网络在线逼近系统未知非线性项,有效解决传统滑模控制中抖振效应与参数不确定性之间的矛盾;最后,将误差作为RBF神经网络的输入,将RBF神经网络的输出作为切换控制,从而实现滑模控制律的在线调整。仿真结果表明:通过与传统的反步滑模控制进行比较,文中方法可有效消除传统滑模控制中切换项引起的抖振问题,使系统具有良好的鲁棒性。更多还原

【Abstract】 A backstepping integral sliding mode control(BISMC) trajectory tracking control method for underactuated autonomous undersea vehicle(AUV) based on radial basis function(RBF) neural network was proposed to address the challenges of difficult horizontal trajectory tracking control and weak anti-interference ability in complex marine environments. Firstly, a kinematic controller was designed by employing the backstepping control method to obtain virtual control laws and actual control inputs. In the dynamic controller, integral sliding mode control was introduced to account for the uncertainty factors and possible external disturbances of the system. Meanwhile, an RBF neural network was adopted to approximate the unknown nonlinear terms of the system online, effectively resolving the contradiction between the chattering effect and parameter uncertainty in traditional sliding mode control. By taking the error as the input of the RBF neural network and using the output of the RBF neural network as the switching control, the online adjustment of the sliding mode control law was achieved. The simulation results show that compared with the traditional BISMC, the proposed method can effectively eliminate the chattering problem caused by the switching terms in traditional sliding mode control, enabling the system to exhibit strong robustness.更多还原