增加行为监控指标

agent_ppo/conf/monitor_builder.py

@@ -98,6 +98,81 @@ def build_monitor():
         )
         .end_panel()
         .end_group()
+        .add_group(
+            group_name="行为指标",
+            group_name_en="behavior",
+        )
+        .add_panel(
+            name="原地撞墙步数",
+            name_en="stuck",
+            type="line",
+        )
+        .add_metric(
+            metrics_name="stuck_count",
+            expr="avg(stuck_count{})",
+        )
+        .add_metric(
+            metrics_name="max_stuck_steps",
+            expr="avg(max_stuck_steps{})",
+        )
+        .end_panel()
+        .add_panel(
+            name="回充逃脱次数",
+            name_en="recharge_escape",
+            type="line",
+        )
+        .add_metric(
+            metrics_name="recharge_escape_count",
+            expr="avg(recharge_escape_count{})",
+        )
+        .end_panel()
+        .add_panel(
+            name="NPC危险接近",
+            name_en="npc_danger",
+            type="line",
+        )
+        .add_metric(
+            metrics_name="npc_close_steps",
+            expr="avg(npc_close_steps{})",
+        )
+        .add_metric(
+            metrics_name="npc_danger_steps",
+            expr="avg(npc_danger_steps{})",
+        )
+        .end_panel()
+        .add_panel(
+            name="失败类型",
+            name_en="failure_type",
+            type="line",
+        )
+        .add_metric(
+            metrics_name="npc_collision",
+            expr="avg(npc_collision{})",
+        )
+        .add_metric(
+            metrics_name="battery_fail",
+            expr="avg(battery_fail{})",
+        )
+        .end_panel()
+        .add_panel(
+            name="得分与充电",
+            name_en="score_charge",
+            type="line",
+        )
+        .add_metric(
+            metrics_name="total_score",
+            expr="avg(total_score{})",
+        )
+        .add_metric(
+            metrics_name="charge_count",
+            expr="avg(charge_count{})",
+        )
+        .add_metric(
+            metrics_name="remaining_charge",
+            expr="avg(remaining_charge{})",
+        )
+        .end_panel()
+        .end_group()
         .build()
     )
     return config_dict

agent_ppo/feature/preprocessor.py

@@ -80,6 +80,9 @@ class Preprocessor:
         self.stuck_steps = 0
         self.recharge_no_progress_steps = 0
         self.fake_charger_steps = 0
+        self.stuck_count = 0
+        self.max_stuck_steps = 0
+        self.recharge_escape_count = 0
 
         self.dirt_cleaned = 0
         self.last_dirt_cleaned = 0
@@ -138,6 +141,10 @@ class Preprocessor:
         self.nearest_npc_dist = float(self.GRID_SIZE)
         self.npc_danger = False
         self.npcs = []
+        self.npc_close_steps = 0
+        self.npc_danger_steps = 0
+        self.npc_collision = 0
+        self.battery_fail = 0
 
         self.local_dirt_ratio = 0.0
         self.local_obstacle_ratio = 0.0
@@ -412,6 +419,8 @@ class Preprocessor:
         if self.prev_pos is not None and self.cur_pos == self.prev_pos and 0 <= self.last_action < 8:
             if self.charge_delta <= 0:
                 self.stuck_steps += 1
+                self.stuck_count += 1
+                self.max_stuck_steps = max(self.max_stuck_steps, self.stuck_steps)
             else:
                 self.stuck_steps = 0
         else:
@@ -431,6 +440,17 @@ class Preprocessor:
         )
         self.recharge_no_progress_steps = self.recharge_no_progress_steps + 1 if no_progress else 0
 
+        if self.step_no > 0 and self.nearest_npc_dist <= 3:
+            self.npc_close_steps += 1
+        if self.step_no > 0 and self.npc_danger:
+            self.npc_danger_steps += 1
+
+        if self.terminated and not self.truncated:
+            if self.battery <= 0 or self.remaining_charge <= 0:
+                self.battery_fail = 1
+            if self.npc_danger or self.nearest_npc_dist <= 1:
+                self.npc_collision = 1
+
     def _need_recharge_escape(self):
         return self.stuck_steps >= 2 or self.recharge_no_progress_steps >= 5 or self.fake_charger_steps >= 2
 
@@ -775,6 +795,7 @@ class Preprocessor:
         if not ranked:
             return list(recharge_action)
 
+        self.recharge_escape_count += 1
         ranked.sort()
         escape = [0] * 8
         for _, action in ranked[: min(4, len(ranked))]:

agent_ppo/workflow/train_workflow.py

@@ -177,6 +177,12 @@ class EpisodeRunner:
                         f"remaining_charge:{remaining_charge} "
                         f"charge_count:{charge_count} "
                         f"recharge_steps:{fm.recharge_steps} "
+                        f"stuck_count:{fm.stuck_count} "
+                        f"max_stuck_steps:{fm.max_stuck_steps} "
+                        f"recharge_escape_count:{fm.recharge_escape_count} "
+                        f"npc_close_steps:{fm.npc_close_steps} "
+                        f"npc_danger_steps:{fm.npc_danger_steps} "
+                        f"npc_collision:{fm.npc_collision} "
                         f"nearest_charger:{fm.nearest_charger_range_dist:.1f} "
                         f"nearest_npc:{fm.nearest_npc_dist:.1f} "
                         f"result_code:{result_code} "
@@ -215,6 +221,16 @@ class EpisodeRunner:
                                 os.getpid(): {
                                     "reward": total_reward + final_reward,
                                     "episode_cnt": self.episode_cnt,
+                                    "total_score": float(total_score),
+                                    "stuck_count": float(fm.stuck_count),
+                                    "max_stuck_steps": float(fm.max_stuck_steps),
+                                    "recharge_escape_count": float(fm.recharge_escape_count),
+                                    "npc_close_steps": float(fm.npc_close_steps),
+                                    "npc_danger_steps": float(fm.npc_danger_steps),
+                                    "npc_collision": float(fm.npc_collision),
+                                    "battery_fail": float(fm.battery_fail),
+                                    "charge_count": float(charge_count),
+                                    "remaining_charge": float(remaining_charge),
                                 }
                             }
                         )