Automated Parameter Tuning for Above-Knee Powered Prostheses

描述（由申请人提供）：传统的下肢假肢装置不能提供步态的净功率，因此限制了截肢者的运动。动力假肢装置有可能为截肢者提供自然运动；然而，对于如何控制这些动力装置，我们的知识还存在差距。在这项研究中，我们提出了一种适用于人体运动控制建模的参数识别技术。我们将使用数学模型来捕捉人类运动的复杂动力学，包括地面撞击。随后，我们将开发一种基于逆最优控制（IOC）方法的参数识别技术。IOC使我们能够计算出最能解释受控动力系统所观察到的行为的代价函数。应用于人类运动的观察（例如，运动捕捉数据），该方法将产生一个成本函数，这反过来将产生假肢参数的最佳选择。因此，该方法可以自动化冗长的参数调整过程，并显着减少下肢截肢者的临床就诊次数和持续时间。

DESCRIPTION (provided by applicant): Traditional lower-limb prosthetic devices do not provide net power over a gait, and therefore limit the locomotion of amputees. Powered prostheses devices have the potential of providing natural locomotion to amputees; however, there is a gap in our knowledge as to how these powered devices should be controlled. In this study we propose to develop a parameter identification technique applicable to human locomotion control modeling. We will use mathematical models to capture the complex dynamics of human locomotion, including ground impacts. Subsequently, we will develop a parameter identification technique based on the method of Inverse Optimal Control (IOC). IOC allows us to compute the cost function that best explains the observed behavior of a controlled dynamical system. Applied to observations of human locomotion (e.g., motion capture data), this method will produce a cost function, which in turn will produce an optimal choice of parameters for the prosthesis. This method can therefore automate the lengthy process of parameter tuning, and significantly reduce the number and duration of clinical visits for lower-limb amputees.