Prosthetics, Exoskeletons, and Assistive Robots

Why are sensors required in biomechatronics?

1. Sense variable in the human subject

   1. Movement related

   2. Physiological

2. Sense environmental variables

   1. Sound (cochlear implant)

Common Sensors

Power Sources

Typically require power for the mechatronics devices

• 12V for motors

• 5V for analogue electronics

• 3.3V for signal processor

These systems may be:

• Static - power should not present problems for design

• Portable - require some form of battery power

Batteries

• Battery power is one of the limiting factors due to their limited life.

• Can be non-rechargeable (primary) or rechargable (secondary).

• Primary batteries generally have a better power density, used for low power applications.

• If access is difficult, secondary batteries should be used.

• Measured in Watt-Hours

Energy Harvesting

The process of converting ambient sources of energy into electrical energy.

Can be harvested from a number of different sources:

• Body motion

• Vibration

• Changes in shape or volume and pressure

• Temperature gradient

• Blood glucose

In current research the devices are not particularly efficient

Challenges in Exoskeletons

• Portability

• Ease of use

• Metabolic cost

• Lightweight actuators

• Actuators that can supply high torque at slow speeds

• Actuators that are soft and flexible

• Intuitive control systems

• Control systems that adapt to the changing dynamic of the human

Soft Exoskeleton Control

• Layered control paradigm

  • Hierarchical control with 3 layers in cascade

• High level controller

  • Understands user intention, converts to estimated joint torque / position using dynamic model of human arm

• Mid level controller

  • Compensate for nonlinear backlash phenomenon

• Low level controller

  • Sends input to DC motor to compensate for nonlinear friction

Models of Dynamic Systems

It is often the case that models are used that transform measured output variables into input variables that are used to drive actuators.

This is done with the inverse model of a system.

For models of physical systems the inverse often has more zeros than poles.

For this case the transfer function magnitude will be unbounded at high frequency (can be solved with a low pass filter)

Linear Dynamic Control

• Control by negative feedback → reject disturbances

• Control by positive feedback → sensitive to disturbances (i.e. BLEEK exoskeleton)

  • Controller designed close to inverse of exoskeleton dynamics

  • Subject to instability if the inverse model is inaccurate