[WIP] Fix target position to navigate to arena center

[Copilot]

Fix VBot Navigation Target Position to Use Three-Stage Navigation

Problem

Current implementation sets target positions as random offsets from robot initial position, causing robots to navigate to random locations instead of the arena trigger points.

Solution Plan

• 1. Update cfg.py to define target_point_a and target_point_b
• 2. Modify reset() method in vbot_section001_np.py to set targets based on dog_stage
• 3. Add logic to update targets when stage changes in update_state() method
• 4. Test the changes to ensure three-stage navigation works correctly

Key Changes

1. Stage 0 → 1: Target = [0.0, 1.5] (inner circle trigger point A) +1 point
2. Stage 1 → 2: Target = [0.0, 0.0] (center trigger point B) +1 point
3. Removed random offset target generation and now use fixed stage-based targets

Changes Made

• Added target_point_a and target_point_b to VBotSection001EnvCfg in cfg.py
• Modified reset() method to initialize all dogs to stage 0 with target = target_point_a
• Added dynamic target update logic in update_state() that changes targets based on dog_stage
• Targets now automatically switch from inner circle to center when dogs reach trigger points

Original prompt

问题描述

VBot 导航竞技场中，导航终点目标位置设定不正确，导致机器狗没有导航到圆心。

根本原因 ❌

在 reset() 方法中，目标位置是相对于机器狗初始位置的随机偏移：

# 当前代码（错误）
cmd_range = cfg.commands.pose_command_range  # [-5.0, -5.0, -3.14, 5.0, 5.0, 3.14]
target_positions = robot_init_pos[:, :2] + sampled[:, :2]  # 相对初始位置的偏移

这导致：

• 机器狗初始位置：外圈随机（3.0m）
• 目标位置：初始位置 + 随机偏移（-5.0~5.0）
• 问题：目标不在圆心，位置完全随机！

要求 ✅

根据题目规范，竞技场导航应该有三阶段目标：

阶段0 → 1: 外圈 → 内圈 (触发点A: [0.0, 1.5])   +1分
阶段1 → 2: 内圈 → 圆心 (触发点B: [0.0, 0.0])   +1分

修复内容

1. 在 cfg.py 中定义竞技场参数（如果未定义）：

   @dataclass
   class VBotSection001EnvCfg:
       # 竞技场参数
       arena_center: list = field(default_factory=lambda: [0.0, 0.0])
       arena_outer_radius: float = 3.0
       arena_inner_radius: float = 1.5
       boundary_radius: float = 3.5
       target_point_a: list = field(default_factory=lambda: [0.0, 1.5])  # 内圈
       target_point_b: list = field(default_factory=lambda: [0.0, 0.0])  # 圆心

2. 修改 vbot_section001_np.py 的 reset() 方法：

   • 将目标位置生成从"相对初始位置的偏移"改为"固定的竞技场触发点"
   • 初始阶段：目标为 target_point_a（内圈）
   • 当机器狗到达 target_point_a 后，目标切换为 target_point_b（圆心）

3. 实现三阶段导航逻辑：

   # 根据当前阶段设置目标位置
   if dog_stage[i] == 0:  # 未到达内圈
       target = target_point_a  # [0.0, 1.5]
   elif dog_stage[i] == 1:  # 已到达内圈，前往圆心
       target = target_point_b  # [0.0, 0.0]

关键文件

• motrix_envs/src/motrix_envs/navigation/vbot/cfg.py - 配置参数
• motrix_envs/src/motrix_envs/navigation/vbot/vbot_section001_np.py - 目标位置生成和阶段管理

参考文档

• ACCEPTANCE_CHECKLIST.md - 竞技场参数规范
• ENVIRONMENT_COMPARISON.md - 三阶段目标定义
• USAGE_GUIDE.md - 任务流程说明

This pull request was created from Copilot chat.

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.