作 者: ;
机构地区: 湖南大学电气与信息工程学院电气工程系
出 处: 《控制理论与应用》 1997年第5期748-753,共6页
摘 要: 本文提出一种神经网络自组织控制器,并应用于非线性跟踪控制中,为了加快神经模糊控制器的在线学习,文中给出了一种变斜率的是速梯度下降学习算法.仿真结果表明,该控制是有效的。 A neural fuzzy logic self-organizing controller is proposed and applied to control of a non-linear learning tracking problem- The proposed controller can self adjust its control rules by modifying the weights of the network on-line such that it can improve response performance of the nonlinear plant. In this paper,a weight-learning algorithm is presented using the gradient descent method with a veriable-slopes,which is useful to accelerate learning and improve convergence,The effectiveness of the proposed control scheme is illustrated through simulations.
领 域: [自动化与计算机技术] [自动化与计算机技术] [理学] [理学]