From 616dd06e785efb182e2db0c3da1afdec422cfcab Mon Sep 17 00:00:00 2001
From: ChenXi <chenxiliu2405@gmail.com>
Date: Sun, 22 Mar 2026 02:32:58 -0400
Subject: [PATCH] Amend success rewards

---
 rl_game/get_up/config/t1_env_cfg.py | 19 +++++++++----------
 1 file changed, 9 insertions(+), 10 deletions(-)

diff --git a/rl_game/get_up/config/t1_env_cfg.py b/rl_game/get_up/config/t1_env_cfg.py
index d39a1ab..8562d50 100644
--- a/rl_game/get_up/config/t1_env_cfg.py
+++ b/rl_game/get_up/config/t1_env_cfg.py
@@ -135,13 +135,6 @@ def is_standing_still(
 
     return env.extras["stable_timer"] > standing_time
 
-def joint_deviation_l2(env: ManagerBasedRLEnv, asset_cfg: SceneEntityCfg) -> torch.Tensor:
-    """计算关节相对于默认姿态(default_joint_pos)的偏差平方和"""
-    # 获取当前关节位置相对于默认位置的差值
-    # mdp.joint_pos_rel 返回的是 (current_pos - default_pos)
-    diff = mdp.joint_pos_rel(env, asset_cfg)
-    return torch.sum(torch.square(diff), dim=-1)
-
 # --- 2. 配置类 ---
 
 T1_JOINT_NAMES = [
@@ -250,9 +243,15 @@ class T1GetUpRewardCfg:
 
     # 6. 成功终极大奖
     is_success = RewTerm(
-        func=lambda env, keys: env.termination_manager.get_term(keys).float(),
-        weight=1000.0,
-        params={"keys": "standing_success"}
+        func=is_standing_still,
+        weight=2000.0,
+        params={
+            "min_head_height": 1.05,
+            "min_pelvis_height": 0.75,
+            "max_angle_error": 0.3,
+            "standing_time": 0.2,
+            "velocity_threshold": 0.5
+        }
     )
 
 @configclass