MME/ECE 436/436H/536: Control of Dynamic Systems
1-1
ECE/MME 436/536:
Control of Dynamic Systems
Lab Project

Prepared by
Dr. James Chagdes

Department of Mechanical and Manufacturing Engineering
College of Engineering and Computing
Miami University

MME/ECE 436/436H/536: Quanser/My-Rio Lab Manual
1-2
1 Position Control of QUBE-Servo
1.1 Introduction
The Quanser QUBE-Servo is composed of a brushed DC motor that is connected to a PWM
amplifier (shown in Figure 1.1).

Figure 1.1: Quanser QUBE-Servo DC motor with inertia disk attachment
The objective of the project is to control the angular position of the disk when exposed to a step
and ramp input. The myRIO-LabVIEW™ interface will allow the controller to be implemented on
the real system.
1.2 Performance Specification
The designed controller should achieve the following performance specifications:
1. When exposed to a step input:
a. The peak time ( pt ) should be 0.2 seconds.
b. The percentage overshoot ( %OS ) should be 20%.

2. When exposed to a ramp input:
a. Zero steady-state error ( % sse ).
1.3 Report Format
1.3.1 Final Report
You must submit a report detailing the entire model and design process. The report must be
comprehensive and include a cover sheet, table of contents, executive summary, a body with
relevant figures and tables placed near the text, and appendices showing all calculations and
relevant work. Schematics of the complete system must also be generated as well as a short
description of its complete operation. The body of the report should consist of three sections:
Modeling, Design (including theoretical proof or justification), and Simulation/Implementation.
MME/ECE 436/436H/536: Control of Dynamic Systems
1-3
Clear and concise writing will result in higher scores. The report must be typed but some
calculations in appendices may be hand written.
1.3.2 Modeling
It is crucial for the success of the project that the cart system model accurately captures all of the
important dynamics of the system. Your careful work during this process will enable you to have
closely matched simulations (in MATLAB) and implementations (on the real system) and give
you much less frustration in meeting the required specifications. On the other hand, you may want
to build a model that is simple enough for controller design purpose. Moreover, there is always
mismatch between a model and the actual plant anyway, and a well-designed feedback controller
could handle the mismatch. Making a good trade-off between these two aspects would ease the
design process quite a lot. The modeling process is open ended; you have the freedom to choose
whatever methods you think are most appropriate.
A good first step in modeling the system might be to draw free body diagrams for each component
of the system and write the corresponding dynamics equations, such as equations of motion.
Certain assumptions can be made to simplify the model as long as you can justify that they are
reasonable. Transfer functions can be obtained after linearization of the dynamic equations. You
are encouraged to develop novel approaches to system identification and to generate multiple
models of your system!
1.3.3 Controller Design
Clearly state the methodology used and underlying assumptions. Analyze the specifications and
determine the controller architecture to achieve the specifications. All design decisions must be
fully supported by an analysis of the original system, the desired system, and any trade-offs
involved in the decision. Controller parameters may be refined through iteration; however, they
must be first estimated by a sound analysis. You are expected not only to develop a controller
that meets the specifications, but also to choose controller parameters for optimum performance.
You have complete freedom in the controller design process, but you must explain and justify your selection.
Briefly describe your entire thought process from start to finish and include that work in an appendix. The
designed controller should be clearly stated in the report, the stability and performance of the
resulting nominal closed-loop system must be theoretically proven or justified.
1.3.4 Implementation/Simulation
Your controller should be tested on your own simulation model first to make sure that the design
is stable and would not damage the equipment. Always simulate your design before
implementing it with hardware!
Then you may implement the design in a LabVIEW VI (downloaded from Canvas) to control the
real system.
You should comment on the differences between the model simulation and the actual
implementation on the system. It’s highly likely that you will have to “tweak” the controller to get
the best performance on the real system. What are the most significant causes for these differences?
How do nonlinearities affect the results? Include simulation and implementation parameters such
MME/ECE 436/436H/536: Quanser/My-Rio Lab Manual
1-4
as sampling rate. Also include a LabVIEW block diagram that shows how your controller was
implemented and include the mathematical model used for the simulation. Both simulated and
implemented responses should meet the minimum specifications.