Intermittent predictive control of man and machine
Lead Research Organisation:
University of Glasgow
Department Name: Mechanical Engineering
Abstract
Abstracts are not currently available in GtR for all funded research. This is normally because the abstract was not required at the time of proposal submission, but may be because it included sensitive information such as personal details.
Publications
Gawthrop P
(2009)
Predictive feedback in human simulated pendulum balancing.
in Biological cybernetics
Gawthrop P
(2013)
Power-constrained intermittent control
in International Journal of Control
Gawthrop P
(2012)
Semi-active damping using a hybrid control approach
in Journal of Intelligent Material Systems and Structures
Gawthrop P
(2014)
Intermittent control models of human standing: similarities and differences.
in Biological cybernetics
Gawthrop P
(2011)
The system-matched hold and the intermittent control separation principle
in International Journal of Control
Gawthrop P
(2013)
Human stick balancing: an intermittent control explanation.
in Biological cybernetics
Gawthrop P
(2009)
Event-driven intermittent control
in International Journal of Control
Gawthrop P
(2012)
Intermittent tapping control
in Proceedings of the Institution of Mechanical Engineers, Part I: Journal of Systems and Control Engineering
Gawthrop P
(2010)
Act-and-Wait and Intermittent Control: Some Comments
in IEEE Transactions on Control Systems Technology
Gawthrop P
(2010)
Intermittent redesign of continuous controllers
in International Journal of Control
Description | This project is part of a three centre project ("Intermittent Predictive Control of Man and Machine" EP/F068514/1, EP/F069022/1 and EP/F06974X/1). Together, our primary aim was to develop intermittent control theory as a new paradigm for engineering and physiological control. Intermittent control uses a sequence of parametrised open-loop control trajectories whose parameters are adjusted intermittently using feedback. We identified three objectives, corresponding to the three project parts: - Control of balance: to provide physiological evidence discriminating intermittent open-loop from continuous feedback mechanisms in human balance. - Control of posture: to provide physiological evidence of intermittent open loop control of posture in general. - System identification and control. Within this project part we are addressing the last Objective: to develop the theoretical basis and system identification tools to discriminate intermittent open loop from continuous feedback control, and to develop and implement intermittent controllers that emulate natural human control, in an engineering context. Key findings are reviewed and summarised in (Gawthrop, Loram, Lakie and Gollee 2011). We have developed descriptions for intermittent open loop control in the time and frequency domains (Gawthrop 2009) which has enabled us to implement system identification methods in both domains, and to evaluate these with simulated data and experimental recordings from humans performing a visual-manual control task (Gollee et al. 2012). This has led to the discovery of the "masquerading property" of intermittent control: for sustained control tasks an intermittent controller can appear indistinguishable from a similarly designed continuous controller, up to a certain bandwidth. This is important as it shows that sustained human control can be equally well explained by intermittent open loop control and by continuous control. This finding provides a theoretical basis for the development of the experimental methods by the project partners to distinguish intermittent from continuous control, by generalising the concept of refractoriness to higher order systems. We have generalised intermittent control with a constant open-loop interval, to event driven intermittent control (Gawthrop and Wang, 2009). We have shown that intermittent control provides an alternative explanation for non-linear remnant which does not require an additional source of noise (Mamma et al. 2011). We have formulated intermittent control to use discretely different structural bases for trajectory construction such as an underlying continuous ("system-matched") or impulsive ("tapping") design, leading to a theoretical basis for tapping control (Gawthrop and Gollee 2012, Loram et al. 2011). In the same context we found that the separation principle remains valid if a system matched hold is used (Gawthrop and Wang 2011). We found that intermittent control provides important advantages for engineering control system, in the context of semi-active damping (Gawthrop, Neild and Wagg 2012), power-constrained vibration control (Gawthrop, Wagg, Neild and Wang 2012), and as a way to overcome friction and other non-linearities (Gawthrop and Gollee 2012). In summary, we have established the theoretical basis of intermittent open loop control and developed system identification tools which can differentiate intermittent from continuous control. We have demonstrated how intermittent control can be used in engineering applications. |
Exploitation Route | We have shown that intermittent control can provide the basis for a range of engineering control problems such as semi-active damping, vibration control and control of application with non-linear actuators. With the outcome of this project, intermittent control can now be readily employed in these and other engineering applications. Intermittent control is inherently variable which can be exploited in the context of adaptive control and learning. As a result, it is highly suitable for applications in robotics, in particular in the context of human machine interaction. Applications in assistive and rehabilitation systems (such as neuroprosthetics and rehabilitation robotics) or are of primary interest, since intermittent control can help to optimise the interaction with the human operator and enhance the assistive and rehabilitative outcome. The results of this research contribute to the understanding of human motor control, which can inform diagnostics and clinical intervention methods in neuromuscular dysfunction. We are exploring how the research findings can be used in the context of clinical applications of control in rehabilitation engineering. In particular, we plan to develop systems, based on intermittent control, which make use of the open-loop interval to enhance robustness and improve rehabilitation outcome when used in neuroprosthetics or rehabilitation robotics. Our primary route to transfer these results into a clinical setting is through our established collaboration with the Queen Elizabeth National Spinal Injuries Unit, where we explore applications in spinal cord injury rehabilitation. The basic concept of intermittent open-loop control marks a departure from traditional engineering control principles which have been based on continuous actuation and sensing. The theoretical framework developed in this project provides the basis for future fundamental new developments in engineering control, adaptation and learning. |
Sectors | Digital/Communication/Information Technologies (including Software) Healthcare |
URL | http://www.icmm.org.uk |
Description | An important aspect of intermittent control is that it can provide an explanation and theoretical framework for adaptation and learning, and for the role which variability plays in this context. We are developing methods to apply this understanding in the context of rehabilitation following neurological injury, e.g. in spinal cord injury, where re-learning of function plays an important role. This leads to better rehabilitation outcome, with associated reduced health-care costs. |
First Year Of Impact | 2013 |
Sector | Healthcare |
Impact Types | Economic |
Description | CONACyT International Scholarships |
Amount | £76,698 (GBP) |
Funding ID | 314429 |
Organisation | National Council on Science and Technology (CONACYT) |
Sector | Public |
Country | Mexico |
Start | 09/2013 |
End | 09/2016 |
Description | Standard proposal |
Amount | £2,998,222 (GBP) |
Funding ID | EP/R018634/1 |
Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
Sector | Public |
Country | United Kingdom |
Start | 04/2018 |
End | 04/2022 |
Description | Multivariable intermittent control for learning and adaptation |
Organisation | Delft University of Technology (TU Delft) |
Department | Robotics Institute |
Country | Netherlands |
Sector | Academic/University |
PI Contribution | - Develop new control engineering approaches to extend intermittent control to adaptation and learning - implement and evaluate multi-variable intermittent control |
Collaborator Contribution | Collaboration in all related areas, including robotics, human movement control, and control theory |
Impact | Multi-disciplinary collaboration, involving engineering, human physiology and robotics. Outputs include software implementations of algorithms and simulation methods. |
Start Year | 2014 |
Description | Multivariable intermittent control for learning and adaptation |
Organisation | Manchester Metropolitan University |
Department | School of Healthcare Science |
Country | United Kingdom |
Sector | Academic/University |
PI Contribution | - Develop new control engineering approaches to extend intermittent control to adaptation and learning - implement and evaluate multi-variable intermittent control |
Collaborator Contribution | Collaboration in all related areas, including robotics, human movement control, and control theory |
Impact | Multi-disciplinary collaboration, involving engineering, human physiology and robotics. Outputs include software implementations of algorithms and simulation methods. |
Start Year | 2014 |
Description | Multivariable intermittent control for learning and adaptation |
Organisation | University of Melbourne |
Country | Australia |
Sector | Academic/University |
PI Contribution | - Develop new control engineering approaches to extend intermittent control to adaptation and learning - implement and evaluate multi-variable intermittent control |
Collaborator Contribution | Collaboration in all related areas, including robotics, human movement control, and control theory |
Impact | Multi-disciplinary collaboration, involving engineering, human physiology and robotics. Outputs include software implementations of algorithms and simulation methods. |
Start Year | 2014 |
Description | robotICs: do autonomous robots benefit from an Intermittent Control (IC) implementation? |
Organisation | Delft University of Technology (TU Delft) |
Department | Robotics Institute |
Country | Netherlands |
Sector | Academic/University |
PI Contribution | - provide control theory knowledge and expertise - 6 months research placement by PhD student in Delft to transfer intermittent control implementation to robotic device |
Collaborator Contribution | - access to robotic systems to implement intermittent control algorithms - expertise in the area of robotics |
Impact | - PhD training - knowledge transfer |
Start Year | 2014 |
Description | robotICs: do autonomous robots benefit from an Intermittent Control (IC) implementation? |
Organisation | Manchester Metropolitan University |
Department | School of Healthcare Science |
Country | United Kingdom |
Sector | Academic/University |
PI Contribution | - provide control theory knowledge and expertise - 6 months research placement by PhD student in Delft to transfer intermittent control implementation to robotic device |
Collaborator Contribution | - access to robotic systems to implement intermittent control algorithms - expertise in the area of robotics |
Impact | - PhD training - knowledge transfer |
Start Year | 2014 |