Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd
Upload
10 views

Programming IV - Advanced Motor Control

Uploaded by

arjunalwe24
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
10 views

Programming IV - Advanced Motor Control

Uploaded by

arjunalwe24
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 24

PROGRAMMING IV

Advanced Motor Control


What are some problems with
controlling motors like this?
Open Loop Control
Closed Loop Control
Open vs Closed Loop
Simpler to set up Complicated setup

Requires little testing Requires tuning

Gets the job done for simple Almost required for complex

systems systems

Requires you to know what Can independently generate

power to set the system to the input


PID Controller
Proportional, Integral, Derivative -> PID

PID Controller ≠ Motor Controller

Mathematical construct that helps a system

reach some state


PID Equation
Setpoint or r(t): The desired state of the system with

respect to time

Output or y(t): The measured state of the system with

respect to time

Error or e(t): The difference between the setpoint and

the output.

Control Effort or u(t): The output added by the

controller with respect to time


How does PID work?
Uses 3 constants: kP, kI, kD (known as gains)

Performs calculations on the error and adds

the result to the input, thereby reducing the

error over time.


Proportional Term

u(t) = kP * e(t)

Makes sure the output reaches the setpoint

Does most of the heavy lifting


Integral Term
u(t) = kI * ∫ e(t)dt

kI multiplies the sum of all error

Eliminates “steady state error”

Usually not needed


Derivative Term
u(t) = kD * d/dt e(t)

kD multiplies the rate of change of e(t)

Tries to make the rate of change of the output

match the rate of change of the setpoint

Helps slow the system down when it

approaches the setpoint


What happens when kP is too high?
A large kP causes overshoot and oscillation around the

setpoint

Can be fixed by lowering kP or increasing kD


What happens when kI is too high?
A large kI can cause overshoot

The Integrator term can result in “wind-up”

If the Integrator isn‘t reset, it can add to much to the

output
What happens when kD is too high?
A large kD can result in unpredicatable behaviour and

aggressively dampen the system


How are gains calculated
PID gains can be determined experimentally

through trial and error, until the system behaves

as expected

WPILib has a tool called SysID that can calculate

gains for you by performing a test routine


Setting up PID
How do you track the current state of a system?

RECALL: Encoders measure rotational movement.

Brushless motors (like Neos) have built-in

encoders)
PID Classes
WPILib PIDController

REV SparkPIDController

CRTE Phoenix Velocity and Position

ControlMode
Adding PID
Create a PIDController object with some gains

Use .calculate to get u(t). .calculate requires y(t)

and r(t) as parameters.


Adding PID

Since y(t) and r(t) are both in units of rotations,

u(t) will be in units of rotations.


.set() vs .setVoltage()

.set() uses a percent value

.setVoltage() uses a voltage value

.setVoltage prevents instability when the battery

voltage drops
How to get RPM for a Neos

Create a RelativeEncoder object by calling

.getEncoder on a SparkMAX.

.getVelocity returns the measured speed.


How to get RPM for a Neos

Create a RelativeEncoder object by calling

.getEncoder on a SparkMAX.

.getVelocity returns the measured speed.


Tips

.calculate is not a set-it and forget-it method, so it

must be called continuously. You may need to

update your Commands to use runEnd instead of

startEnd.

You might also like