中文摘要：

步态训练轨迹是影响康复训练效果的一项重要因素，而自适应性对于下肢康复机器人的临床应用具有重要的意义．振荡器可通过在线调节参数而输出不同波形的周期信号，常用于康复机器人步态轨迹的生成．本文在高斯核函数非线性振荡器的基础上提出了一种下肢康复机器人步态轨迹自适应算法．该算法通过轨迹偏差实现对参考轨迹波形的调节，并且用相位偏差曲线面积实现参考轨迹周期的自适应．本文首先介绍了用于生成步态参考轨迹的非线性振荡器的数学模型；其次，详细描述了基于该模型的参考轨迹波形和周期自适应算法；最后，以悬挂减重式下肢康复机器人为研究对象，建立机器人与人体下肢仿真模型，对所提出的步态参考轨迹自适应算法进行仿真实验，并验证了该算法的可行性．

英文摘要：

Gait trajectory is an important factor to the effects of rehabilitation with adaptability has great significance for clinical application of lower limb rehabilitation robot. The oscillator can output various waveforms and periodic smooth signals by adjusting parameters online, which is frequently used to generate gait reference trajectory for rehabilitation robots. This paper proposes a gait trajectory adaptation algorithm for lower limb rehabilitation robot based on Gaussian kernel nonlinear oscillators. In the algorithm, trajectory deviation is used to adjust the waveform of reference trajectory, and the area of phase deviation curve is used to adjust the period of reference trajectory. Firstly, the paper introduces the nonlinear oscillator model for generating gait reference trajectory. Then, the adaptation algorithms for gait trajectory waveform and period are described based on this mathematical model. Finally, a human-robot simulation model is built based on a weight supporting rehabilitation robot system. These algorithms are validated by simulation experiments and the results demonstrate the feasibility of proposed algorithms.