Unify ROS params: replace C++ PARAM_ macros with YAML-based config structs (vision filter + planning)

[Copilot]

Description

Replaces the custom DEFINE_*/DECLARE_*/PARAM_ macro-based parameter system with plain config structs populated from YAML via this->get_parameter(). This eliminates the indirection of global static variables and aligns with standard ROS2 parameter patterns.

Vision Filter module (~33 files):

• New VisionFilterConfig struct in params.hpp holds all vision filter parameters
• VisionFilter node loads params from YAML in load_config(), passes config down the class hierarchy: World → Camera → KalmanBall/KalmanRobot → KalmanFilter2D/3D, kick detectors, bounce, world ball/robot
• Removed all DEFINE_NS_FLOAT64, DECLARE_*, using PARAM_* from vision filter sources
• Removed global_params.hpp and LocalROS2ParamProvider dependencies from vision filter

Planning module (~28 files):

• New PlanningConfig struct with all planner parameters (constraints, RRT, replanner, collect, settle, escape, pivot, intermediate)
• PlannerNode loads config and threads it through PlannerForRobot → PathPlanner subclasses via PlanRequest
• Updated MotionConstraints/RotationConstraints to use hardcoded defaults instead of PARAM_ references
• Updated all path planners: collect, settle, escape, pivot, line_pivot, rotate, plus replanner, create_path, rrt_util

Example — before:

#include <rj_param_utils/vision/vision_params.hpp>
DEFINE_NS_FLOAT64(kVisionFilterParamModule, ball, init_covariance, 100.0, "...")
// usage:
double p = ball::PARAM_init_covariance;

Example — after:

// VisionFilter node:
this->get_parameter("ball.init_covariance", cfg.ball.init_covariance);
// Non-node class receives config in constructor:
double p = config_.ball.init_covariance;

Not yet addressed (follow-up work):

• Control module (motion_control.cpp)
• Common module (world_state.cpp, packet_convert.cpp)
• Radio and joystick modules
• Removal of now-unused rj_param_utils infrastructure (param.hpp macros, global_params.cpp/hpp, planning_params.cpp/hpp, vision_params.cpp/hpp)
• YAML file consolidation

Associated Issue

Unify ROS params across all files for consistent YAML-based access.

Design Documents

Pattern follows ROS2 Global Parameters — nodes load params from YAML, pass values to non-node classes via config structs.

Steps to test

Test Case 1

1. Build with make
2. Run vision filter and planning tests
3. Launch soccer.launch.py with sim_params.yaml — verify vision filter and planner initialize with correct parameter values

Expected result: All parameters load from YAML identically to previous macro-based behavior. No functional change.

(Optional) Sub-issues (for drafts)

• Vision filter module: config struct + all source/test files updated
• Planning module: config struct + all planner/primitive files updated
• Control module (motion_control)
• Common module (world_state, packet_convert)
• Radio + joystick modules
• Remove unused rj_param_utils macro infrastructure
• Consolidate/deduplicate YAML parameter files

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.