论文部分内容阅读
讨论了载体位置不受控、姿态受控情况下,自由漂浮柔性空间机械臂的高斯基模糊神经网络自学习控制问题。利用拉格朗日方程和模态综合法可建立柔性空间机械臂的动力学模型,但由于此类空间机器人系统严格遵守动量守恒,其动力学方程表现出强烈的非线性性质。结合神经网络和模糊控制,即利用神经网络来实现模糊推理可使模糊控制具有自学习能力,在此基础上,设计了柔性空间机械臂关节空间的高斯基模糊神经网络自学习控制方案。由于将动量守恒定理耦合到系统动力学方程的推导过程中,所提出的控制方案具有不需要测量、反馈载体位置、移动速度和移动加速度的显著优点。系统的数值仿真,证实了方法的有效性。
The Gaussian fuzzy neural network self-learning control problem of free-floating flexible space manipulator is discussed under the condition of uncontrolled position and controlled attitude. The dynamic model of flexible space manipulator can be established by using Lagrange equation and modal synthesis method. However, due to the strict observance of the conservation of momentum, the dynamic equations of this space robot exhibit strong nonlinearity. Combining neural network and fuzzy control, that is, using neural network to realize fuzzy inference, fuzzy control can be self-learning. Based on this, Gaussian fuzzy neural network self-learning control scheme of flexible space manipulator joint space is designed. Due to the coupling of the law of conservation of momentum into the derivation of the system dynamics equations, the proposed control scheme has the significant advantage of not requiring measurement, feedback carrier position, moving speed and moving acceleration. The numerical simulation of the system confirms the validity of the method.