- Page 3 Assuming That You Want To Help A Patient Who Has A Stroke Regain The Mobility Of The Arms The First Task You Wan 1 (57.38 KiB) Viewed 28 times
Page 3 Assuming that you want to help a patient who has a stroke regain the mobility of the arms. The first task you want to do is to drive (regulate) his right arm to a certain position (angle) of course, this task can be easily achieved if you apply an appropriate torque. However, you can neither directly apply a torque to the joint nor know the appropriate torque. You can apply a very low-level current to stimulate the nerves around the patient's muscles to cause muscle activation, and then the muscle activation generates torques. Fortunately, you designed an excellent controller. You want to use this controller to provide appropriate torques (if it is not appropriate the controller can adjust it). Unfortunately, your controller needs the knowledge of the full state, which includes the angular position, angular velocity, and muscle activation. If your sensor cannot measure muscle activation, you cannot use the controller. Let's think about what to do. 1g. How do you want to design your sensor? What states it should be able to measure? How do you want to estimate muscle activation? (Assuming that the arm dynamics can be linearized using the Jacobian matrices) We can see that there is only one muscle getting stimulated. You can only change the angle to one direction (you cannot apply a negative current to make move to the other direction). In this case, how do you want to design the control gains? (Hint: Do you want oscillation for the controlled system?) Controller Sensor n Stimulator