Project Overview

About the Project

This project allows the user to experiment with PID control to match the speed to two DC motors on a robot platform. A webservice hosted on the mBed allows the user to connect to the robot over WiFi, using a webclient such as Firefox, to change PID gains and setpoint.

Parts List and Vendors

Pin Connection Tables

Code

The program code can be imported from here:

Webpage and microSD Card

The webpage is loaded into the mBed from the microSD card. You will need to copy and paste the webpage HTML text into a HTML file and save to the SD card with the given file name.

PID Control

This project uses motor shaft encoders to provide feedback to a PID controller to set the speed to the given setpoint.

Introduction/Overview of PID Controls

Proportional, Integral, Derivative (PID) Control is one of the most common control schemes used in industry. Arguably, the greatest feature of PID control is its robustness. This robustness arises form the fact that the controller does not have to have a completely accurate plant model, in fact we can know little about the plant and tune the controller to the plant through trial and error. For an introduction to PID see Brian Douglas's Introduction to PID https://www.youtube.com/watch?v=UR0hOmjaHp0 and also Brian's PID Example's https://www.youtube.com/watch?v=XfAt6hNV8XM

Tuning PID Controller

Many papers and methods have been developed on the topic of PID tuning. There is no perfect method for tuning that has yet been developed, and tuning remains largely empirical. A quick web search on the topic will produce many results on the subject, simple to advanced. For quick and dirty approach the steps below can be followed:

• Set all gains to zero
• Increase the Proportional gain at small increments until suitable rise time is achieved with little oscillations (acceptable overshoot is application dependent).
• Once the Proportional gain is set apply the same approach to the Integral gain. The Integral gain should be selected such it eliminates small errors that cannot be overcome by the Proportional gain. The Integral gain should be such that it does not introduce undamped oscillations and maintains steady- state tracking.
• Lastly, once the Proportional and Integral gains are selected, use the same approach to select the Derivative gain. The Derivative gain seeks to damp overshoot and oscillations to reach steady- state tracking more quickly.

About the Code

The PID controller code is based on Brett Beaugard's Arduino PID Library. An in depth explanation of the PID code development can be found here: http://brettbeauregard.com/blog/2011/04/improving-the-beginners-pid-introduction/

Networking

About the Code

The webserver code is based on the project here: https://developer.mbed.org/users/4180_1/notebook/using-the-esp8266-with-the-mbed-lpc1768/. Two methods in this new project handle receiving and parsing user input and updating the webpage, these are "parseInput()" and "updateWebpage()". These two methods are custom to the webpage. Should you desire to change the webpage at all, i.e. add new user input/output blocks, you will need to modify these methods.The code is well commented and serves as a template.

Results

Open- Loop Control

When in open- loop mode the PID controllers are turned off and no feedback is used in setting speed of the motor. Because no two motors are exactly the same one may turn at a different rate, even with the same applied voltage/ PWM Duty Cycle. This causes the robot to travel in a curve path.

Closed- Loop Control with PID

By using PID control and matching the setpoints of each motor, feedback is used by the PID controllers which command the output to the setpoint. By doing this we can match the speeds of both motors resulting in straight line travel.