|
|
|
![]() |
|
|||||||
|
||||||||
![]() |
|
|
Thread Tools | Rate Thread | Display Modes |
|
|
|
#1
|
||||
|
||||
|
Cross-Language Question ... Control Systems Practices ... Encoders/Gyros/Accels
I've been struggling with some decisions for a while now and have finally decided to surrender to the "cloud" of knowledge & advice here on CD.
In practice, in general, do you find yourself depending on these various sensors for different reasons within your code? We have many navigational sensors at our disposal and sometimes I find it hard to decide which one to pick. When do you choose to use the encoder values over the gyro angles? When do you trust your accels over your encoder positions? I find myself jumping between encoder distances (but then worry if/when we lose traction), and the gyro (which drifts, but is good for short bursts), and the accels that summarize the situation (provided we haven't been recently jarred). For you advanced control systems gurus, what sorts of advice can you offer for all of these various trade-offs. I'm interested in general overall theories/practices ... and ... to perhaps spur some conversation, provide a simple example of the best way to "perfectly" rotate your 'bot by exactly 90-degrees (CW or CCW) with the single push of a joystick button. There may be subtle differences depending on the drive-train type, but I'd like to steer the conversation towards more "generic" theory/practice. ... I'm all ears ... ![]() |
|
#2
|
||||
|
||||
|
Re: Cross-Language Question ... Control Systems Practices ... Encoders/Gyros/Accels
For the past three year's I've coded our robot, we've alway stuck to encoders for autonomous naviagation (albeit to a much more limited extent last year due to all the sliding). We've played around with the idea of using gyros and accelerometers, but their drift over time and sensitivity to bumping has always led us to distrust them.
|
|
#3
|
||||
|
||||
|
Re: Cross-Language Question ... Control Systems Practices ... Encoders/Gyros/Accels
Our team uses the gyro to drive straight. It may seem unnecessary, but a little difference in the speed of the motors and/or gearboxes results in the robot curving when driven straight. The gyro, despite it's drift, is capable of detecting the unwanted rotation and, after a little scaling, the error between that rotation and the desired rotation can simply be added to the motors. You'd be amazed how much this makes a difference in driving.
As for encoders, we generally use them for controlling any mechanism that is driven by a motor and requires control feedback to operate. |
![]() |
| Thread Tools | |
| Display Modes | Rate This Thread |
|
|
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| Question: Team practices to promote/support academics | evergreen | General Forum | 13 | 31-08-2008 23:47 |
| Wiring Gyros and Encoders? | SuspectZero | Electrical | 3 | 03-02-2008 15:30 |
| Position idlers to get better chain wrap -- a question about good practices. | Madison | Technical Discussion | 23 | 31-01-2008 09:35 |
| encoders/gyros/potentiometers | nehalita | Programming | 12 | 31-01-2005 23:08 |
| Control Systems | archiver | 2000 | 1 | 23-06-2002 22:15 |