Iterative Learning Control with Human-in-the-loop

人机交互的迭代学习控制

• 批准号: 2482288
• 金额: --
• 依托单位: University of Southampton
• 依托单位国家: 英国
• 项目类别: Studentship
• 财政年份: 2020
• 资助国家: 英国
• 起止时间: 2020 至 无数据
• 项目状态: 已结题
• 来源: https://gtr.ukri.org/projects?ref=studentship-2482288
• 关键词: Iterative; Learning; Control; Human; loop

This project will develop a new type of controller that will significantly improve the effectiveness of stroke rehabilitation technologies. The framework will be first to simultaneously assist the user while embedding a model of their own neurological motor control system within the system. This will allow it to provide effective assistance and therapy which responds to their voluntary intention and residual movement capabilities. For example, the human user may be trying to reach their arm out to pick up an object, and the control system must support their intention while maximising their contribution to the movement. The approach will involve a substantial theoretical development of "Iterative Learning Control" (ILC) which uses the idea of learning from experience to improve tracking control, and is applied to systems that perform the same task several types. The theory will involve components of adaption, learning and robust performance design and will be tested experimentally by applying functional electrical stimulation to help users move their muscles. This component will involve a small-scale clinical trial with stroke participants facilitated by the second supervisor. This project will produce results that are ground-breaking for the development of new more effective rehabilitation and assistive technologies.

该项目将开发一种新型控制器，将显着提高中风康复技术的有效性。该框架将首先帮助用户，同时在系统中嵌入他们自己的神经运动控制系统模型。这将使它能够提供有效的援助和治疗，以回应他们的自愿意愿和剩余的运动能力。例如，人类用户可能试图伸出手臂来拿起物体，控制系统必须支持他们的意图，同时最大限度地提高他们对运动的贡献。该方法将涉及“迭代学习控制”（ILC）的大量理论发展，它使用从经验中学习的思想来改善跟踪控制，并应用于执行相同任务的几种类型的系统。该理论将涉及适应，学习和强大的性能设计的组成部分，并将通过应用功能性电刺激来帮助用户移动他们的肌肉进行实验测试。该部分将涉及一项小规模临床试验，由第二名主管协助卒中参与者。该项目将产生突破性的成果，以开发新的更有效的康复和辅助技术。

项目成果

Model Predictive Valve Control of Lung Pressure Profile Tracking

肺压力分布跟踪的模型预测阀控制

• DOI: 10.1109/anzcc56036.2022.9966865
• 发表时间: 2022
• 作者: Thompson M