仿蜘蛛救援机械臂结构设计与研究

Design and Research on the Structure of the Spider Rescue Robot Arm

针对非结构化环境下的救援任务,结合生物界蜘蛛腿部的运动特性,设计了一种仿蜘蛛式救援机械臂。基于机构学数学基础的拉格朗日方程求解救援机械臂的动力学模型。鉴于救援机器人的实时性、自适应等特性,分析径向基函数神经网络的控制策略,提出基于径向基函数神经网络救援机械臂的控制策略。采用径向基三层前向网络设计救援机器人的控制方案,重点探讨了救援机器人单关节的神经网络控制方法,并对其进行Matlab仿真。结果表明:所设计的控制方法可实现救援机械臂的自适应性和鲁棒性。

In view of the rescue mission in the unstructured environment,a kind of artificial spider type rescue robot research platform is designed,combined with the motion characteristics of biological spider legs. The dynamic model of that rescue robot is solved based on the Lagrange equation of mechanism mathematical. And,the control strategy based on radial basis function neural network of rescue robot platform is put forward,in view of the characteristics of rescue robot,such as real-time and adaptive. Then,the control scheme of rescue robot based on RBF three layer feed forward is studied,focusing on neural network control method to probe the rescue robot joint. Finally,the control scheme is simulated in the Matlab software,showing that the robustness and self-adaptiveness of the robot is received by that control method.