⚙️ Configuration Management System (dconfig)

The configuration management system provides centralized parameter management, supporting multi-robot configuration and runtime hot reload.

📋 System Overview

Core functions of the dconfig module:

Centralized Configuration: Unified management of all module parameters
Multi-robot Support: Independent configuration for each robot
Runtime Hot Reload: Update parameters without restarting
Configuration Monitoring: Automatic detection of configuration file changes

📁 Configuration File Structure

RoboCup_Workspace/core/src/dconfig/
├── constant.yml          # Field constants (shared by all robots)
├── global.yml            # Global configuration
├── 1/                    # Robot 1 configuration
│   ├── dvision.yaml      # Vision parameters
│   ├── dmotion.yaml      # Motion parameters
│   └── dancer_io.yaml    # Hardware parameters
├── 2/                    # Robot 2 configuration
│   ├── dvision.yaml
│   ├── dmotion.yaml
│   └── dancer_io.yaml
├── 3/                    # Robot 3 configuration
├── 4/                    # Robot 4 configuration
├── 5/                    # Robot 5 configuration
└── 6/                    # Robot 6 configuration

🏟️ Field Constants Configuration

constant.yml

Defines field geometry parameters, shared by all robots.

yaml

dbehavior:
  geometry:
    # Field dimensions (unit: centimeters)
    field_length: 900        # Field length
    field_width: 600         # Field width

    # Goal parameters
    goal_depth: 60           # Goal depth
    goal_width: 260          # Goal width
    goal_height: 180         # Goal height

    # Penalty area
    goal_area_length: 20     # Goal area length
    goal_area_width: 300     # Goal area width
    penalty_mark_distance: 150  # Penalty mark distance

    # Field markers
    center_circle_diameter: 150  # Center circle diameter
    border_strip_width: 70       # Border strip width
    line_width: 5                # Line width

  robot:
    # Robot parameters
    num_player: 6            # Team size
    max_pitch: 70            # Max pitch angle
    min_pitch: -15           # Min pitch angle
    max_yaw: 75              # Max yaw angle
    min_yaw: -75             # Min yaw angle
    walk_speed: 450 / 26     # Walking speed
    turn_speed: 10           # Turning speed

  team:
    teaminfo_outdated: 0.5   # Team info expiry time
    ball_share_enabled: False  # Enable ball sharing
    pace: False              # Enable pace control

🌍 Global Configuration

global.yml

Defines game-related global parameters.

yaml

ZJUDancer:
  # Simulation mode
  Simulation: false
  OfflineRecord: false

  # Robot settings
  RobotID: 1              # Robot number (1-6)
  Role: GCStriker         # Role: Striker/Defender/GoalKeeper (GC prefix = use GameController)
  AttackRight: True       # True = attack right goal

  # GameController settings
  UseGameController: true
  GameControllerAddress: "192.168.1.100"
  TeamNumber: 17

  # Debug settings
  Debug: false
  LogLevel: "INFO"

Role Configuration

Available roles: Striker, Defender, GoalKeeper Add GC prefix to listen to referee box commands, e.g. GCStriker

AttackRight

UseGameController and GC prefix must be changed together.

🤖 Per-Robot Configuration

Vision Parameters (dvision.yaml)

yaml

dvision:
  camera:
    device: "/dev/video0"
    width: 640
    height: 480
    fps: 30

  # Camera intrinsics (unique per robot, requires calibration)
  camera_matrix:
    fx: 615.123
    fy: 615.123
    cx: 320.0
    cy: 240.0

  distortion:
    k1: -0.123
    k2: 0.045
    p1: 0.001
    p2: -0.002

  # Detection parameters
  ball_detection:
    confidence_threshold: 0.7
    min_radius: 10
    max_radius: 100

Motion Parameters (dmotion.yaml)

yaml

dmotion:
  walk:
    max_speed: 0.3          # Max walking speed (m/s)
    acceleration: 0.1       # Acceleration
    step_height: 0.03       # Foot lift height

  balance:
    kp: 0.5
    ki: 0.1
    kd: 0.05

  kick:
    power_scale: 1.0        # Kick power multiplier
    direction_offset: 0.0   # Direction correction offset

🔄 Runtime Hot Reload

Modifying Parameters at Runtime

bash

# Modify a single parameter
rosservice call /dconfig/set_param \
  '{key: "dbehavior.striker.kick_distance", value: "0.2"}'

# Reload entire configuration file
rosservice call /dconfig/reload_file \
  '{file: "global.yml"}'

# Reload all configurations
rosservice call /dconfig/reload_all '{}'

Listen to Configuration Changes

python

import rospy
from dconfig.msg import ConfigUpdate

def on_config_update(msg):
    print(f"Config updated: {msg.key} = {msg.value}")
    # Update local variables
    if msg.key == "striker.kick_distance":
        self.kick_distance = float(msg.value)

rospy.Subscriber('/dconfig/updates', ConfigUpdate, on_config_update)

📊 Role Starting Position Configuration

role.yml

Defines each role's starting position on the field.

yaml

# dbehavior/config/role.yml
Striker:
  start_pos: [-100, -200]    # Starting position (cm, field coordinates)
  kickoff_pos: [-50, 0]      # Position after entering field

Defender:
  start_pos: [-100, 200]
  kickoff_pos: [-200, 0]

GoalKeeper:
  start_pos: [-420, 0]
  kickoff_pos: [-420, 0]

Coordinate System

Robot faces the positive Y-axis when entering. Right is positive X-axis. Origin is field center. Coordinates assume own half is left half (AttackRight=True). When AttackRight=True, start_pos is mirrored about Y-axis.

🛠️ Configuration Best Practices

1. Per-Robot Calibration

Each robot requires independent camera calibration:

bash

# Camera calibration
rosrun dvision camera_calibration --robot_id 1

# IMU calibration
rosrun dancer_io imu_calibration --robot_id 1

# Motion parameter tuning
rosrun dmotion motion_tuning --robot_id 1

2. Configuration Backup

Regularly back up configuration files:

bash

# Backup config
tar -czf config_backup_$(date +%Y%m%d).tar.gz dconfig/

# Restore config
tar -xzf config_backup_20260313.tar.gz

3. Version Control

Use Git to manage configuration files:

bash

cd RoboCup_Workspace/core/src/dconfig
git add .
git commit -m "Update robot 1 vision parameters"
git push

Vision System - Vision parameter configuration
Behavior Decision - Behavior parameter configuration
Motion Control - Motion parameter configuration
Complete Tutorial - Configuration in practice

💡 FAQ

Q: How to create a configuration for a new robot? A: Copy an existing robot configuration directory and modify robot-specific parameters (e.g., camera calibration).

Q: Configuration changes not taking effect, what should I do? A: Check if config_watchdog is running, manually restart the relevant module.

Q: How to share configuration between multiple robots? A: Put common parameters in constant.yml and global.yml; put robot-specific parameters in each robot's directory.

Q: Configuration file format error, what should I do? A: Use the validate_config tool to check syntax and ensure correct YAML format.

⚙️ Configuration Management System (dconfig) ​

📋 System Overview ​

📁 Configuration File Structure ​

🏟️ Field Constants Configuration ​

constant.yml ​

🌍 Global Configuration ​

global.yml ​

🤖 Per-Robot Configuration ​

Vision Parameters (dvision.yaml) ​

Motion Parameters (dmotion.yaml) ​

🔄 Runtime Hot Reload ​

Modifying Parameters at Runtime ​

Listen to Configuration Changes ​

📊 Role Starting Position Configuration ​

role.yml ​

🛠️ Configuration Best Practices ​

1. Per-Robot Calibration ​

2. Configuration Backup ​

3. Version Control ​

🔗 Related Documentation ​

💡 FAQ ​

⚙️ Configuration Management System (dconfig)

📋 System Overview

📁 Configuration File Structure

🏟️ Field Constants Configuration

constant.yml

🌍 Global Configuration

global.yml

🤖 Per-Robot Configuration

Vision Parameters (dvision.yaml)

Motion Parameters (dmotion.yaml)

🔄 Runtime Hot Reload

Modifying Parameters at Runtime

Listen to Configuration Changes

📊 Role Starting Position Configuration

role.yml

🛠️ Configuration Best Practices

1. Per-Robot Calibration

2. Configuration Backup

3. Version Control

🔗 Related Documentation

💡 FAQ