基于小波网络的下肢携行外骨骼系统模型辨识研究

Carrying Lower Extreme Exoskeleton Model Identification Based on Wavelet Neural Network

采用虚拟样机技术对外骨骼进行建模,同时应用虚拟样机模型对系统动态方程的逆模型采用小波网络进行学习逼近,充分利用了小波网络在时域及频域所具有的非线性映射能力。理论分析及仿真结果证明了此模型逼近方法的可行性及有效性。

The virtual prototyping technology is introduced to model the exoskeleton. Using the prototype model,the wavelet neural network is applied to dynamic exoskeleton model study,which takes full advantages of its good time-frequency nonlinear mapping. Theoretical analysis and simulation results test the feasibility and validity of this model learning method.