摘要
为了提高助行康复机器人在不同坡度上的稳定性,合理的步态控制方法显得尤为重要。这里以助行康复机器人为研究对象,提出了一种将长短期记忆网络和动态运动基元结合的步行康复机器人自适应步态控制方法。将传感器采集信息与动态运动基元中步态控制项相结合,对参考步态进行调整。同时通过更改动态运动基元的时间项和空间项对步态进行二次调整以适应不同的坡度。通过试验对该控制方法的可用性进行验证。结果表明,该方法可以在(0~20)°斜坡进行自适应步态变换,具有一定的实用价值。
In order to improve the stability of walking rehabilitation robot on different slopes,a reasonable gait control method is particularly important.It takes the Walking aid recovery robot as a research object,an adaptive gait control method of walking re-habilitation robot based on short-term memory network and dynamic motion primitives is proposed in this paper.The reference gait is adjusted by combining the information collected by the sensor with the gait control items in the dynamic motion primitive.At the same time,the gait is adjusted twice by changing the time and space items of the dynamic motion primitive to adapt to dif-ferent slopes.The availability of the control method is verified by experiments.The results show that this method can carry out adaptive gait transformation on(0~20)°degree slope,it has certain practical value.
作者
罗莎
罗思思
卢运娇
王成
LUO Sha;LUO Si-si;LU Yun-jiao;WANG Cheng(Beihai Vocational College,Guangxi Beihai 536000,China;Hu’nan University,Hu’nan Changsha 410082,China)
出处
《机械设计与制造》
北大核心
2024年第6期362-366,共5页
Machinery Design & Manufacture
基金
广西省职业教育教学改革研究项目(GXGZJG2016A008)。
关键词
机器人
步态控制
长短期记忆网络
动态运动基元
助行
自适应
Robot
Gait Control
Long Short Term Memory
Dynamic Movement Primitives
Walking Aid
Self-Adaption
