Scholarlink

<p>Soft robots are often powered by soft actuators, and the dielectric elastomer actuator (DEA) is widely recognized as a prospective soft actuators. Nevertheless, DEAs have complex nonlinear properties, which pose a formidable obstacle for their control. In this paper, a dynamic model of the DEA is established based on Koopman theory and the extended dynamic model decomposition method. This dynamic model is a low-dimensional linear model, which is simple and easy to implement in practical control. Then, to reduce the influence of model uncertainty and external disturbances on control accuracy, a single-neuron adaptive control method and a radial basis function (RBF) network-based adaptive control method are proposed for tracking control of the DEA. Next, to prevent damage to the DEA during the process of controller parameter tuning, a two-stage method is presented to tune the parameters of the single-neuron adaptive controller and RBF network-based adaptive controller. Finally, the efficiency of the proposed control methods is verified via actual tracking control experiments with three differential trajectories.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Dynamic Modeling and Adaptive Control of Dielectric Elastomer Actuators Based on Koopman Theory

  • 期刊论文

  • 发表日期：

    2026年3月14日

查看全文
  • Yawu Wang,
  • Yue Zhang,
  • Dun Mao,
  • Chun-Yi Su

摘要

Soft robots are often powered by soft actuators, and the dielectric elastomer actuator (DEA) is widely recognized as a prospective soft actuators. Nevertheless, DEAs have complex nonlinear properties, which pose a formidable obstacle for their control. In this paper, a dynamic model of the DEA is established based on Koopman theory and the extended dynamic model decomposition method. This dynamic model is a low-dimensional linear model, which is simple and easy to implement in practical control. Then, to reduce the influence of model uncertainty and external disturbances on control accuracy, a single-neuron adaptive control method and a radial basis function (RBF) network-based adaptive control method are proposed for tracking control of the DEA. Next, to prevent damage to the DEA during the process of controller parameter tuning, a two-stage method is presented to tune the parameters of the single-neuron adaptive controller and RBF network-based adaptive controller. Finally, the efficiency of the proposed control methods is verified via actual tracking control experiments with three differential trajectories.