拟人机器人的建模

为解决机构复杂的拟人机器人运动学和动力学问题，提出了基于传统机理结合神经网络的建模方法．首先，从几何构造分析出拟人机器人各部分运动规律．其次，采用Lagrange—Euler方法结合对各部分交互作用力的定义得到动力学模型，交互力用神经网络来实现．最后，将此模型对应于实际的机器人参数，通过仿真证明了建模的有效性．结果表明此方法实用、有效．

Dynamic Modeling and Analysis for a Humanoid Robot

To resolve the kinematics modeling and dynamics modeling of a humanoid robot with complicated machine, a modeling method based on the conventional mechanism combined with neural network is presented. First the kinematics formulas are deduced by analyzing its geometry structure. Second, the dynamics models are built up with Lagrange-Euler method combined with the define of interactive force among all parts of the humanoid robot. The interactive force is realized by neural network method. Finally, the true robot parameters are applied to the model. The simulation results show the validity of the modeling. So, the method is practical and available.