FTC RobotController(SDK) v6.0发布

FTC RobotController(SDK) v6.0发布

by 北一 张 -
Number of replies: 0

Version 6.0 (20200921-085816)

Important Notes

  • Version 6.0 is the minimum legal version for the Ultimate Goal season.
  • Prebuilt APKs are now only available from the repository's Releases page.
  • Android Studio 4.0 required.
  • Android Studio users need to be connected to the internet the first time they build the app (in order to download needed packages for the build). Clearing gradle caches will also require an internet connection for the next build.
  • Version 5.5 was a moderately large off-season, August 2020, drop. It's worth reviewing those release notes below also.
  • Version 5.5 and greater will not work on older Android 4.x and 5.x phones. Users must upgrade to an approved Android 6.x device or newer.
  • The default PIDF values for REV motors have been reverted to the default PID values that were used in the 2018-2019 season
    • This change was made because the 2018-2019 values turned out to work better for many mechanisms
    • This brings the behavior of the REV motors in line with the behavior of all other motors
    • If you prefer the 2019-2020 season's behavior for REV motors, here are the PIDF values that were in place, so that you can manually set them in your OpModes:

      HD Hex motors (all gearboxes):
      Velocity PIDF values: P = 1.17I = 0.117F = 11.7
      Position PIDF values: P = 5.0
      Core Hex motor:
      Velocity PIDF values: P = 4.96I = 0.496F = 49.6
      Position PIDF values: P = 5.0

New features

  • Includes TensorFlow inference model and sample op modes to detect Ultimate Goal Starter Stacks (four rings vs single ring stack).
  • Includes Vuforia Ultimate Goal vision targets and sample op modes.
  • Introduces a digital zoom feature for TensorFlow object detection (to detect objects more accurately at greater distances).
  • Adds configuration entry for the REV UltraPlanetary HD Hex motor

Enhancements

  • Adds setGain() and getGain() methods to the NormalizedColorSensor interface
    • By setting the gain of a color sensor, you can adjust for different lighting conditions. For example, if you detect lower color values than expected, you can increase the gain.
    • The gain value is only applied to the argb() and getNormalizedColors() methods, not to the raw color methods. The getNormalizedColors() method is recommended for ease-of-use and clarity, since argb() has to be converted.
    • Updates SensorColor Java sample to demonstrate gain usage
  • Merges SensorREVColorDistance Java sample into SensorColor Java sample, which showcases best practices for all color sensors
  • Improves retrieving values from the REV Color Sensor V3
    • Updates the normalization calculation of the RGB channels
    • Improves the calculation of the alpha channel (can be used as an overall brightness indicator)
    • Fixes the default sensor resolution, which caused issues with bright environments
    • Adds support for changing the resolution and measuring rate of the Broadcom sensor chip
    • Removes IR readings and calculations not meant for the Broadcom sensor chip

Bug fixes

  • Improved reliability of BNO055IMU IMU initialization to prevent random initialization failures (which manifested as Problem with 'imu').