Chapter 5 HW 2 Hints
Chapter 5 HW 2 Hints
Chapter 5 HW 2 Hints
Chapter 5 HW 2 Hints
Problem 1. This is an unstable plant whose linear transfer results from the linearization of the inverse-square magnetic law. (a) Be careful when you factor the denominator of the transfer function: the roots are not complex. (b) The discretization is pretty standard. (c) I used a lead compensator to meet the specications; I used the zero of the lead compensator to cancel one of the plant poles (not the unstable onenever cancel unstable poles!) (d) Introduce reference input xr in the standard manner, and nd the two desired step responses. I decided to use Simulink to simulate this system, since I could examine both the ball position x and the amplier current i using the same model.
Problem 2. You can discretize the plant in the usual way and nd the z -plane regions that are acceptable to place the poles. (a) You can actually meet these specications using proportional control D(z ) = Kp . (b) Redraw the block diagram to show grade Gr as the input and velocity V as the output. The sensor, sampler, compensator D(z ), ZOH, throttle dynamics, and fuel dynamics will all migrate to the feedback path. Youll need to discretize the forward path (vehicle dynamics) and the feedback path (all the aforementioned stu) separately to get an eective G(z ) and H (z )...then apply the grade step input of 3 and see how much speed output you get. This speed output is the error. Ideally a grade disturbance should cause no speed output (this is disturbance rejection). My controller had a speed error of a little over 1 mph for the 3% grade disturbance. (c) To reduce the error to zero, you need to increase the system Type. This is done by adding an integrator, so at the least you will need PI control. However, you do not need PID control. If youve ever driven a car with a cruise control that hunts you know the desirability of a cruise control with critical damping or even overdamping.