Making a Duck Car
Systems Engineering - Negative Feedback Controllers
In Control Theory class, my partner and I designed and built an analog electronic proportional-derivative (PD) motor control system to make an RC ‘duck car’ that maintains a certain distance between it and the object in front of it, like a baby and mother duck. Overall, the car showed rapid and well damped response which met the specifications for rise time, settling time, overshoot, and steady-state error.
Characterization
Firstly we characterized the car – how quickly it accelerated to maximum velocity, and what the input-output relationship of the proximity sensor was. After this we ‘closed the loop’ to show what the uncompensated negative feedback system looks like (pictured above).
Compensator Design
Once we’d characterized our system, we designed a compensator to generate a well damped, responsive, and low error system. We achieved this using a host of control theory principles, the SISOtool MATLAB modeling package for root-locus based design, and experimental troubleshooting. We then built our compensator from OpAmps to establish our first version.
Testing and Refinement
We refined our compensatory by experimental tweaking of the system to establish a duck car with no overshoot, sub 1 second settling time, and an error of about 1% after 2 seconds. This project taught us the methods and knowledge necessary to build control systems which can be applied to more complex systems with applications in nearly every engineering discipline.