Control Augmentation Design for Some Nonlinear Dynamic Systems with Human-in-the-loop
Saved in:
| Title: | Control Augmentation Design for Some Nonlinear Dynamic Systems with Human-in-the-loop |
|---|---|
| Authors: | Dai, Andong |
| Committee Members: | Xu, Yunjun |
| Summary: | Control augmentation is a control strategy that has been wildly studied in human-in-the-loop systems. A human control model is crucial to the performance of such control augmentation since it allows the controller to act accordingly via predicting the human operator's control behavior. The benefit of the control augmentation is limited by the fact that the human control models are inaccurate in general. In this study, two types of uncertainties, human vision distortion and internal vehicle model distortion are investigated. Based on the proposed models, a Hidden Markov Model based control augmentation framework is studied to assist human operators to control dynamic systems to precisely follow the desired commands. The proposed method is first studied for a truck in row alignment scenarios. Later, the method is improved for a general class of dynamic human-in-the-loop systems. A new general form of the human internal vehicle model is proposed to describe the operator's understanding about the dynamics. A recursive, closed-form solution of the internal vehicle model parameters is derived for a class of dynamical systems so that the computational cost is significantly reduced. |
| URL: | https://stars.library.ucf.edu/etd2020/845 |
| Database: | OpenDissertations |
| Abstract: | Control augmentation is a control strategy that has been wildly studied in human-in-the-loop systems. A human control model is crucial to the performance of such control augmentation since it allows the controller to act accordingly via predicting the human operator's control behavior. The benefit of the control augmentation is limited by the fact that the human control models are inaccurate in general. In this study, two types of uncertainties, human vision distortion and internal vehicle model distortion are investigated. Based on the proposed models, a Hidden Markov Model based control augmentation framework is studied to assist human operators to control dynamic systems to precisely follow the desired commands. The proposed method is first studied for a truck in row alignment scenarios. Later, the method is improved for a general class of dynamic human-in-the-loop systems. A new general form of the human internal vehicle model is proposed to describe the operator's understanding about the dynamics. A recursive, closed-form solution of the internal vehicle model parameters is derived for a class of dynamical systems so that the computational cost is significantly reduced. |
|---|
