fix(scada): use project-scoped metadata

app/algorithms/__init__.py

@@ -1,36 +1,45 @@
-from app.algorithms.cleaning import flow_data_clean, pressure_data_clean
+"""Algorithm package with side-effect-free, lazy compatibility exports."""
-from app.algorithms.sensor import (
-    pressure_sensor_placement_sensitivity,
-    pressure_sensor_placement_kmeans,
-)
-from app.algorithms.isolation.valve import valve_isolation_analysis
-from app.algorithms.leakage import LeakageIdentifier
-from app.algorithms.health import PipelineHealthAnalyzer
-from app.algorithms.burst_location import run_burst_location
-from app.algorithms.simulation.scenarios import (
-    convert_to_local_unit,
-    burst_analysis,
-    valve_close_analysis,
-    flushing_analysis,
-    contaminant_simulation,
-    age_analysis,
-    pressure_regulation,
-)
 
-__all__ = [
+from importlib import import_module
-    "flow_data_clean",
+from typing import Any
-    "pressure_data_clean",
+
-    "pressure_sensor_placement_sensitivity",
+
-    "pressure_sensor_placement_kmeans",
+_EXPORT_MODULES = {
-    "convert_to_local_unit",
+    "flow_data_clean": "app.algorithms.cleaning",
-    "burst_analysis",
+    "pressure_data_clean": "app.algorithms.cleaning",
-    "valve_close_analysis",
+    "pressure_sensor_placement_sensitivity": "app.algorithms.sensor",
-    "flushing_analysis",
+    "pressure_sensor_placement_kmeans": "app.algorithms.sensor",
-    "contaminant_simulation",
+    "valve_isolation_analysis": "app.algorithms.isolation.valve",
-    "age_analysis",
+    "LeakageIdentifier": "app.algorithms.leakage",
-    "pressure_regulation",
+    "PipelineHealthAnalyzer": "app.algorithms.health",
-    "valve_isolation_analysis",
+    "run_burst_location": "app.algorithms.burst_location",
-    "LeakageIdentifier",
+    **{
-    "PipelineHealthAnalyzer",
+        name: "app.algorithms.simulation.scenarios"
-    "run_burst_location",
+        for name in (
-]
+            "convert_to_local_unit",
+            "burst_analysis",
+            "valve_close_analysis",
+            "flushing_analysis",
+            "contaminant_simulation",
+            "age_analysis",
+            "pressure_regulation",
+        )
+    },
+}
+
+__all__ = list(_EXPORT_MODULES)
+
+
+def __getattr__(name: str) -> Any:
+    try:
+        module_name = _EXPORT_MODULES[name]
+    except KeyError as exc:
+        raise AttributeError(f"module {__name__!r} has no attribute {name!r}") from exc
+
+    value = getattr(import_module(module_name), name)
+    globals()[name] = value
+    return value
+
+
+def __dir__() -> list[str]:
+    return sorted({*globals(), *__all__})

app/api/v1/endpoints/project_data.py

@@ -1,10 +1,9 @@
 from fastapi import APIRouter, Depends, HTTPException, Path, Query
 from psycopg import AsyncConnection
 
-import app.native.wndb as wndb
+from app.infra.db.postgresql.scada import ScadaInfoRepository
 from app.infra.db.postgresql.scheme import SchemeRepository
 from app.auth.project_dependencies import get_project_pg_connection
-from app.services import project_info
 
 router = APIRouter()
 
@@ -26,9 +25,7 @@ async def get_scada_info_with_connection(
     返回项目中所有的SCADA设备信息
     """
     try:
-        _ = conn
+        scada_data = await ScadaInfoRepository.get_scadas(conn)
-        network_name = project_info.name
-        scada_data = wndb.get_all_scada_info(network_name) if network_name else []
         return {"success": True, "data": scada_data, "count": len(scada_data)}
     except Exception as e:
         raise HTTPException(

app/infra/db/postgresql/scada.py

@@ -0,0 +1,51 @@
+from typing import Any
+
+from psycopg import AsyncConnection
+
+
+def _optional_text(value: Any) -> str | None:
+    return str(value).strip() if value is not None else None
+
+
+def _optional_float(value: Any) -> float | None:
+    return float(value) if value is not None else None
+
+
+class ScadaInfoRepository:
+    """Read SCADA metadata from the current project's business database."""
+
+    @staticmethod
+    async def get_scadas(conn: AsyncConnection) -> list[dict[str, Any]]:
+        async with conn.cursor() as cur:
+            await cur.execute(
+                """
+                SELECT id,
+                       type,
+                       associated_element_id,
+                       api_query_id,
+                       transmission_mode,
+                       transmission_frequency,
+                       reliability,
+                       x_coor,
+                       y_coor
+                FROM public.scada_info
+                """
+            )
+            records = await cur.fetchall()
+
+        return [
+            {
+                "id": str(record["id"]).strip(),
+                "type": str(record["type"]).strip().lower(),
+                "associated_element_id": _optional_text(
+                    record["associated_element_id"]
+                ),
+                "api_query_id": record["api_query_id"],
+                "transmission_mode": record["transmission_mode"],
+                "transmission_frequency": record["transmission_frequency"],
+                "reliability": _optional_float(record["reliability"]),
+                "x": _optional_float(record["x_coor"]),
+                "y": _optional_float(record["y_coor"]),
+            }
+            for record in records
+        ]

app/infra/db/timescaledb/composite_queries.py

@@ -1,18 +1,19 @@
 import time
-from typing import List, Optional, Any, Dict, Tuple
 from datetime import datetime, timedelta
-from psycopg import AsyncConnection
+from typing import Any, Dict, List, Optional, Tuple
-import pandas as pd
+
 import numpy as np
+import pandas as pd
+from psycopg import AsyncConnection
+
+import app.native.wndb as wndb
 from app.algorithms.cleaning.flow import clean_flow_data_df_kf
 from app.algorithms.cleaning.pressure import clean_pressure_data_df_km
 from app.algorithms.health.analyzer import PipelineHealthAnalyzer
-import app.native.wndb as wndb
+from app.infra.db.postgresql.scada import ScadaInfoRepository
-
 from app.infra.db.timescaledb.repositories.realtime import RealtimeRepository
 from app.infra.db.timescaledb.repositories.scheme import SchemeRepository
 from app.infra.db.timescaledb.repositories.scada import ScadaRepository
-from app.services import project_info
 
 
 class CompositeQueries:
@@ -20,6 +21,13 @@ class CompositeQueries:
     复合查询类，提供跨表查询功能
     """
 
+    @staticmethod
+    async def _get_project_scada_index(
+        postgres_conn: AsyncConnection,
+    ) -> Dict[str, Dict[str, Any]]:
+        scadas = await ScadaInfoRepository.get_scadas(postgres_conn)
+        return {scada["id"]: scada for scada in scadas}
+
     @staticmethod
     async def get_scada_associated_realtime_simulation_data(
         timescale_conn: AsyncConnection,
@@ -48,31 +56,22 @@ class CompositeQueries:
             ValueError: 当 SCADA 设备未找到或字段无效时
         """
         result = {}
-        # 1. 查询所有 SCADA 信息
+        scada_by_id = await CompositeQueries._get_project_scada_index(postgres_conn)
-        network_name = project_info.name
-        scada_infos = wndb.get_all_scada_info(network_name) if network_name else []
 
         for device_id in device_ids:
-            # 2. 根据 device_id 找到对应的 SCADA 信息
+            target_scada = scada_by_id.get(device_id)
-            target_scada = None
-            for scada in scada_infos:
-                if scada["id"] == device_id:
-                    target_scada = scada
-                    break
-
             if not target_scada:
                 raise ValueError(f"SCADA device {device_id} not found")
 
-            # 3. 根据 type 和 associated_element_id 查询对应的模拟数据
             element_id = target_scada["associated_element_id"]
             scada_type = target_scada["type"]
 
-            if scada_type.lower() == "pipe_flow":
+            if scada_type == "pipe_flow":
                 # 查询 link 模拟数据
                 res = await RealtimeRepository.get_link_field_by_time_range(
                     timescale_conn, start_time, end_time, element_id, "flow"
                 )
-            elif scada_type.lower() == "pressure":
+            elif scada_type == "pressure":
                 # 查询 node 模拟数据
                 res = await RealtimeRepository.get_node_field_by_time_range(
                     timescale_conn, start_time, end_time, element_id, "pressure"
@@ -115,26 +114,17 @@ class CompositeQueries:
             ValueError: 当 SCADA 设备未找到或字段无效时
         """
         result = {}
-        # 1. 查询所有 SCADA 信息
+        scada_by_id = await CompositeQueries._get_project_scada_index(postgres_conn)
-        network_name = project_info.name
-        scada_infos = wndb.get_all_scada_info(network_name) if network_name else []
 
         for device_id in device_ids:
-            # 2. 根据 device_id 找到对应的 SCADA 信息
+            target_scada = scada_by_id.get(device_id)
-            target_scada = None
-            for scada in scada_infos:
-                if scada["id"] == device_id:
-                    target_scada = scada
-                    break
-
             if not target_scada:
                 raise ValueError(f"SCADA device {device_id} not found")
 
-            # 3. 根据 type 和 associated_element_id 查询对应的模拟数据
             element_id = target_scada["associated_element_id"]
             scada_type = target_scada["type"]
 
-            if scada_type.lower() == "pipe_flow":
+            if scada_type == "pipe_flow":
                 # 查询 link 模拟数据
                 res = await SchemeRepository.get_link_field_by_scheme_and_time_range(
                     timescale_conn,
@@ -145,7 +135,7 @@ class CompositeQueries:
                     element_id,
                     "flow",
                 )
-            elif scada_type.lower() == "pressure":
+            elif scada_type == "pressure":
                 # 查询 node 模拟数据
                 res = await SchemeRepository.get_node_field_by_scheme_and_time_range(
                     timescale_conn,
@@ -167,19 +157,19 @@ class CompositeQueries:
     @staticmethod
     async def get_realtime_simulation_data(
         timescale_conn: AsyncConnection,
-        featureInfos: List[Tuple[str, str]],
+        feature_infos: List[Tuple[str, str]],
         start_time: datetime,
         end_time: datetime,
     ) -> Dict[str, List[Dict[str, Any]]]:
         """
         获取 link/node 模拟值
 
-        根据传入的 featureInfos，找到关联的 link/node，
+        根据传入的 feature_infos，找到关联的 link/node，
         并根据对应的 type，查询对应的模拟数据
 
         Args:
             timescale_conn: TimescaleDB 异步连接
-            featureInfos: 传入的 feature 信息列表，包含 (element_id, type)
+            feature_infos: 传入的 feature 信息列表，包含 (element_id, type)
             start_time: 开始时间
             end_time: 结束时间
 
@@ -190,20 +180,20 @@ class CompositeQueries:
             ValueError: 当 SCADA 设备未找到或字段无效时
         """
         result = {}
-        for feature_id, type in featureInfos:
+        for feature_id, feature_type in feature_infos:
 
-            if type.lower() == "pipe":
+            if feature_type.lower() == "pipe":
                 # 查询 link 模拟数据
                 res = await RealtimeRepository.get_link_field_by_time_range(
                     timescale_conn, start_time, end_time, feature_id, "flow"
                 )
-            elif type.lower() == "junction":
+            elif feature_type.lower() == "junction":
                 # 查询 node 模拟数据
                 res = await RealtimeRepository.get_node_field_by_time_range(
                     timescale_conn, start_time, end_time, feature_id, "pressure"
                 )
             else:
-                raise ValueError(f"Unknown type: {type}")
+                raise ValueError(f"Unknown type: {feature_type}")
             # 添加 scada_id 到每个数据项
             for item in res:
                 item["feature_id"] = feature_id
@@ -213,7 +203,7 @@ class CompositeQueries:
     @staticmethod
     async def get_scheme_simulation_data(
         timescale_conn: AsyncConnection,
-        featureInfos: List[Tuple[str, str]],
+        feature_infos: List[Tuple[str, str]],
         start_time: datetime,
         end_time: datetime,
         scheme_type: str,
@@ -222,12 +212,12 @@ class CompositeQueries:
         """
         获取 link/node scheme 模拟值
 
-        根据传入的 featureInfos，找到关联的 link/node，
+        根据传入的 feature_infos，找到关联的 link/node，
         并根据对应的 type，查询对应的模拟数据
 
         Args:
             timescale_conn: TimescaleDB 异步连接
-            featureInfos: 传入的 feature 信息列表，包含 (element_id, type)
+            feature_infos: 传入的 feature 信息列表，包含 (element_id, type)
             start_time: 开始时间
             end_time: 结束时间
             scheme_type: 工况类型
@@ -240,8 +230,8 @@ class CompositeQueries:
             ValueError: 当类型无效时
         """
         result = {}
-        for feature_id, type in featureInfos:
+        for feature_id, feature_type in feature_infos:
-            if type.lower() == "pipe":
+            if feature_type.lower() == "pipe":
                 # 查询 link 模拟数据
                 res = await SchemeRepository.get_link_field_by_scheme_and_time_range(
                     timescale_conn,
@@ -252,7 +242,7 @@ class CompositeQueries:
                     feature_id,
                     "flow",
                 )
-            elif type.lower() == "junction":
+            elif feature_type.lower() == "junction":
                 # 查询 node 模拟数据
                 res = await SchemeRepository.get_node_field_by_scheme_and_time_range(
                     timescale_conn,
@@ -264,7 +254,7 @@ class CompositeQueries:
                     "pressure",
                 )
             else:
-                raise ValueError(f"Unknown type: {type}")
+                raise ValueError(f"Unknown type: {feature_type}")
             # 添加 feature_id 到每个数据项
             for item in res:
                 item["feature_id"] = feature_id
@@ -301,33 +291,27 @@ class CompositeQueries:
             ValueError: 当元素类型无效时
         """
 
-        # 1. 查询所有 SCADA 信息
+        scada_by_id = await CompositeQueries._get_project_scada_index(postgres_conn)
-        network_name = project_info.name
+        associated_scada = next(
-        scada_infos = wndb.get_all_scada_info(network_name) if network_name else []
+            (
-
+                scada
-        # 2. 根据 element_type 和 element_id 找到关联的 SCADA 设备
+                for scada in scada_by_id.values()
-        associated_scada = None
+                if scada["associated_element_id"] == element_id
-        for scada in scada_infos:
+            ),
-            if scada["associated_element_id"] == element_id:
+            None,
-                associated_scada = scada
+        )
-                break
 
         if not associated_scada:
-            # 没有找到关联的 SCADA 设备
             return None
 
-        # 3. 通过 SCADA device_id 获取监测数据
         device_id = associated_scada["id"]
 
-        # 根据 use_cleaned 参数选择字段
         data_field = "cleaned_value" if use_cleaned else "monitored_value"
 
-        # 保证 device_id 以列表形式传递
         res = await ScadaRepository.get_scada_field_by_id_time_range(
             timescale_conn, [device_id], start_time, end_time, data_field
         )
 
-        # 将 device_id 替换为 element_id 返回
         return {element_id: res.get(device_id, [])}
 
     @staticmethod
@@ -351,108 +335,124 @@ class CompositeQueries:
             end_time: 结束时间
 
         Returns:
-            "success" 或错误信息
+            "success"
+
+        Raises:
+            ValueError: 当前项目没有可清洗设备或指定时间范围内没有监测数据
         """
-        try:
+        scada_by_id = await CompositeQueries._get_project_scada_index(postgres_conn)
-            # 获取所有 SCADA 信息
+        supported_types = {"pressure", "pipe_flow", "flow"}
-            network_name = project_info.name
-            scada_infos = wndb.get_all_scada_info(network_name) if network_name else []
-            # 将列表转换为字典，以 device_id 为键
-            scada_device_info_dict = {info["id"]: info for info in scada_infos}
 
-            # 如果 device_ids 为空，则处理所有 SCADA 设备
+        if device_ids:
-            if not device_ids:
+            device_ids = [str(device_id).strip() for device_id in device_ids]
-                device_ids = list(scada_device_info_dict.keys())
+            missing_metadata_ids = [
+                device_id
+                for device_id in device_ids
+                if device_id not in scada_by_id
+            ]
+            if missing_metadata_ids:
+                raise ValueError(
+                    f"当前项目中有 {len(missing_metadata_ids)} 个 SCADA 设备缺少元数据"
+                )
 
-            # 批量查询所有设备的数据
+            unsupported_ids = [
-            data = await ScadaRepository.get_scada_field_by_id_time_range(
+                device_id
-                timescale_conn, device_ids, start_time, end_time, "monitored_value"
+                for device_id in device_ids
+                if scada_by_id[device_id]["type"] not in supported_types
+            ]
+            if unsupported_ids:
+                raise ValueError(
+                    f"当前项目中有 {len(unsupported_ids)} 个 SCADA 设备类型不支持清洗"
+                )
+        else:
+            device_ids = [
+                device_id
+                for device_id, info in scada_by_id.items()
+                if info["type"] in supported_types
+            ]
+
+        if not device_ids:
+            raise ValueError("当前项目没有可清洗的 SCADA 设备")
+
+        data = await ScadaRepository.get_scada_field_by_id_time_range(
+            timescale_conn, device_ids, start_time, end_time, "monitored_value"
+        )
+        if not data:
+            raise ValueError("指定时间范围内没有 SCADA 监测数据")
+
+        normalized_data = {
+            str(device_id): records for device_id, records in data.items()
+        }
+        missing_data_ids = [
+            device_id for device_id in device_ids if not normalized_data.get(device_id)
+        ]
+        if missing_data_ids:
+            raise ValueError(
+                f"指定时间范围内有 {len(missing_data_ids)} 个 SCADA 设备没有监测数据"
             )
 
-            if not data:
+        all_records = [
-                return "error: fetch none scada data"  # 没有数据，直接返回
+            {
+                "time": record["time"],
+                "device_id": device_id,
+                "value": record["value"],
+            }
+            for device_id, records in normalized_data.items()
+            for record in records
+        ]
+        if not all_records:
+            raise ValueError("指定时间范围内没有 SCADA 监测数据")
 
-            # 将嵌套字典转换为 DataFrame，使用 time 作为索引
+        df_long = pd.DataFrame(all_records)
-            # data 格式: {device_id: [{"time": "...", "value": ...}, ...]}
+        df = df_long.pivot(index="time", columns="device_id", values="value")
-            all_records = []
+
-            for device_id, records in data.items():
+        pressure_ids = [
-                for record in records:
+            device_id
-                    all_records.append(
+            for device_id in df.columns
-                        {
+            if scada_by_id[device_id]["type"] == "pressure"
-                            "time": record["time"],
+        ]
-                            "device_id": device_id,
+        flow_ids = [
-                            "value": record["value"],
+            device_id
-                        }
+            for device_id in df.columns
+            if scada_by_id[device_id]["type"] in {"pipe_flow", "flow"}
+        ]
+
+        updated_rows = 0
+        for grouped_ids, cleaning_function in (
+            (pressure_ids, clean_pressure_data_df_km),
+            (flow_ids, clean_flow_data_df_kf),
+        ):
+            if not grouped_ids:
+                continue
+
+            source_df = df[grouped_ids].reset_index()
+            cleaned_df = cleaning_function(source_df)
+            time_values = cleaned_df["time"].tolist()
+
+            for device_id in grouped_ids:
+                if device_id not in cleaned_df.columns:
+                    raise ValueError(f"设备 {device_id} 的清洗结果缺少数据列")
+
+                cleaned_values = cleaned_df[device_id].tolist()
+                for time_value, value in zip(time_values, cleaned_values):
+                    time_dt = (
+                        time_value
+                        if isinstance(time_value, datetime)
+                        else datetime.fromisoformat(str(time_value))
                     )
+                    await ScadaRepository.update_scada_field(
+                        timescale_conn,
+                        time_dt,
+                        device_id,
+                        "cleaned_value",
+                        value,
+                    )
+                    updated_rows += 1
 
-            if not all_records:
+        if updated_rows == 0:
-                return "error: fetch none scada data"  # 没有数据，直接返回
+            raise ValueError("SCADA 数据清洗未产生任何数据库更新")
 
-            # 创建 DataFrame 并透视，使 device_id 成为列
+        return "success"
-            df_long = pd.DataFrame(all_records)
-            df = df_long.pivot(index="time", columns="device_id", values="value")
-
-            # 根据type分类设备
-            pressure_ids = [
-                id
-                for id in df.columns
-                if scada_device_info_dict.get(id, {}).get("type") == "pressure"
-            ]
-            flow_ids = [
-                id
-                for id in df.columns
-                if scada_device_info_dict.get(id, {}).get("type") == "pipe_flow"
-            ]
-
-            # 处理pressure数据
-            if pressure_ids:
-                pressure_df = df[pressure_ids]
-                # 重置索引，将 time 变为普通列
-                pressure_df = pressure_df.reset_index()
-                # 调用清洗方法
-                cleaned_df = clean_pressure_data_df_km(pressure_df)
-                # 将清洗后的数据写回数据库
-                for device_id in pressure_ids:
-                    if device_id in cleaned_df.columns:
-                        cleaned_values = cleaned_df[device_id].tolist()
-                        time_values = cleaned_df["time"].tolist()
-                        for i, time_str in enumerate(time_values):
-                            time_dt = datetime.fromisoformat(time_str)
-                            value = cleaned_values[i]
-                            await ScadaRepository.update_scada_field(
-                                timescale_conn,
-                                time_dt,
-                                device_id,
-                                "cleaned_value",
-                                value,
-                            )
-
-            # 处理flow数据
-            if flow_ids:
-                flow_df = df[flow_ids]
-                # 重置索引，将 time 变为普通列
-                flow_df = flow_df.reset_index()
-                # 调用清洗方法
-                cleaned_df = clean_flow_data_df_kf(flow_df)
-                # 将清洗后的数据写回数据库
-                for device_id in flow_ids:
-                    if device_id in cleaned_df.columns:
-                        cleaned_values = cleaned_df[device_id].tolist()
-                        time_values = cleaned_df["time"].tolist()
-                        for i, time_str in enumerate(time_values):
-                            time_dt = datetime.fromisoformat(time_str)
-                            value = cleaned_values[i]
-                            await ScadaRepository.update_scada_field(
-                                timescale_conn,
-                                time_dt,
-                                device_id,
-                                "cleaned_value",
-                                value,
-                            )
-
-            return "success"
-        except Exception as e:
-            return f"error: {str(e)}"
 
     @staticmethod
     async def predict_pipeline_health(

tests/unit/test_postgres_scada_repository.py

@@ -0,0 +1,62 @@
+import asyncio
+
+from app.infra.db.postgresql.scada import ScadaInfoRepository
+
+
+class _FakeCursor:
+    def __init__(self):
+        self.query = None
+
+    async def __aenter__(self):
+        return self
+
+    async def __aexit__(self, exc_type, exc, tb):
+        return False
+
+    async def execute(self, query):
+        self.query = query
+
+    async def fetchall(self):
+        return [
+            {
+                "id": " 25470001 ",
+                "type": " PRESSURE ",
+                "associated_element_id": " J1 ",
+                "api_query_id": "query-1",
+                "transmission_mode": "realtime",
+                "transmission_frequency": None,
+                "reliability": "0.95",
+                "x_coor": "117.1",
+                "y_coor": "32.9",
+            }
+        ]
+
+
+class _FakeConnection:
+    def __init__(self):
+        self.cursor_instance = _FakeCursor()
+
+    def cursor(self):
+        return self.cursor_instance
+
+
+def test_get_scadas_normalizes_id_and_type():
+    conn = _FakeConnection()
+
+    result = asyncio.run(ScadaInfoRepository.get_scadas(conn))
+
+    assert result == [
+        {
+            "id": "25470001",
+            "type": "pressure",
+            "associated_element_id": "J1",
+            "api_query_id": "query-1",
+            "transmission_mode": "realtime",
+            "transmission_frequency": None,
+            "reliability": 0.95,
+            "x": 117.1,
+            "y": 32.9,
+        }
+    ]
+    assert "associated_element_id" in conn.cursor_instance.query
+    assert "FROM public.scada_info" in conn.cursor_instance.query

tests/unit/test_scada_cleaning.py

@@ -0,0 +1,267 @@
+import asyncio
+from datetime import datetime, timezone
+from unittest.mock import AsyncMock
+
+import pandas as pd
+import pytest
+
+from app.api.v1.endpoints import project_data
+from app.infra.db.timescaledb import composite_queries
+
+
+PROJECT_SCADA = {
+    "id": "fengyang-pressure-1",
+    "type": "pressure",
+    "associated_element_id": "J1",
+    "api_query_id": "query-1",
+    "transmission_mode": "realtime",
+    "transmission_frequency": None,
+    "reliability": 1.0,
+    "x": 117.1,
+    "y": 32.9,
+}
+START_TIME = datetime(2026, 6, 1, tzinfo=timezone.utc)
+END_TIME = datetime(2026, 6, 2, tzinfo=timezone.utc)
+
+
+def _cleaning_args():
+    return (
+        object(),
+        object(),
+        ["fengyang-pressure-1"],
+        datetime(2026, 6, 1, tzinfo=timezone.utc),
+        datetime(2026, 6, 2, tzinfo=timezone.utc),
+    )
+
+
+def test_clean_scada_uses_current_project_metadata(monkeypatch):
+    monkeypatch.setattr(
+        composite_queries.ScadaInfoRepository,
+        "get_scadas",
+        AsyncMock(
+            return_value=[{"id": "fengyang-pressure-1", "type": "pressure"}]
+        ),
+    )
+    monkeypatch.setattr(
+        composite_queries.ScadaRepository,
+        "get_scada_field_by_id_time_range",
+        AsyncMock(
+            return_value={
+                "fengyang-pressure-1": [
+                    {"time": "2026-06-01T00:00:00+08:00", "value": 26.5}
+                ]
+            }
+        ),
+    )
+    update_mock = AsyncMock()
+    monkeypatch.setattr(
+        composite_queries.ScadaRepository,
+        "update_scada_field",
+        update_mock,
+    )
+    monkeypatch.setattr(
+        composite_queries,
+        "clean_pressure_data_df_km",
+        lambda frame: pd.DataFrame(
+            {
+                "time": frame["time"],
+                "fengyang-pressure-1": frame["fengyang-pressure-1"],
+            }
+        ),
+    )
+
+    result = asyncio.run(
+        composite_queries.CompositeQueries.clean_scada_data(*_cleaning_args())
+    )
+
+    assert result == "success"
+    update_mock.assert_awaited_once()
+
+
+def test_clean_scada_rejects_devices_missing_from_project_metadata(monkeypatch):
+    monkeypatch.setattr(
+        composite_queries.ScadaInfoRepository,
+        "get_scadas",
+        AsyncMock(return_value=[{"id": "other-device", "type": "pressure"}]),
+    )
+    query_mock = AsyncMock()
+    monkeypatch.setattr(
+        composite_queries.ScadaRepository,
+        "get_scada_field_by_id_time_range",
+        query_mock,
+    )
+
+    with pytest.raises(ValueError, match="缺少元数据"):
+        asyncio.run(
+            composite_queries.CompositeQueries.clean_scada_data(*_cleaning_args())
+        )
+
+    query_mock.assert_not_awaited()
+
+
+def test_clean_scada_rejects_zero_database_updates(monkeypatch):
+    monkeypatch.setattr(
+        composite_queries.ScadaInfoRepository,
+        "get_scadas",
+        AsyncMock(
+            return_value=[{"id": "fengyang-pressure-1", "type": "pressure"}]
+        ),
+    )
+    monkeypatch.setattr(
+        composite_queries.ScadaRepository,
+        "get_scada_field_by_id_time_range",
+        AsyncMock(
+            return_value={
+                "fengyang-pressure-1": [
+                    {"time": "2026-06-01T00:00:00+08:00", "value": 26.5}
+                ]
+            }
+        ),
+    )
+    update_mock = AsyncMock()
+    monkeypatch.setattr(
+        composite_queries.ScadaRepository,
+        "update_scada_field",
+        update_mock,
+    )
+    monkeypatch.setattr(
+        composite_queries,
+        "clean_pressure_data_df_km",
+        lambda _frame: pd.DataFrame(
+            {"time": [], "fengyang-pressure-1": []}
+        ),
+    )
+
+    with pytest.raises(ValueError, match="未产生任何数据库更新"):
+        asyncio.run(
+            composite_queries.CompositeQueries.clean_scada_data(*_cleaning_args())
+        )
+
+    update_mock.assert_not_awaited()
+
+
+def test_clean_scada_propagates_write_failures(monkeypatch):
+    monkeypatch.setattr(
+        composite_queries.ScadaInfoRepository,
+        "get_scadas",
+        AsyncMock(
+            return_value=[{"id": "fengyang-pressure-1", "type": "pressure"}]
+        ),
+    )
+    monkeypatch.setattr(
+        composite_queries.ScadaRepository,
+        "get_scada_field_by_id_time_range",
+        AsyncMock(
+            return_value={
+                "fengyang-pressure-1": [
+                    {"time": "2026-06-01T00:00:00+08:00", "value": 26.5}
+                ]
+            }
+        ),
+    )
+    monkeypatch.setattr(
+        composite_queries.ScadaRepository,
+        "update_scada_field",
+        AsyncMock(side_effect=RuntimeError("database write failed")),
+    )
+    monkeypatch.setattr(
+        composite_queries,
+        "clean_pressure_data_df_km",
+        lambda frame: frame,
+    )
+
+    with pytest.raises(RuntimeError, match="database write failed"):
+        asyncio.run(
+            composite_queries.CompositeQueries.clean_scada_data(*_cleaning_args())
+        )
+
+
+def _patch_project_scada_metadata(monkeypatch):
+    monkeypatch.setattr(
+        composite_queries.ScadaInfoRepository,
+        "get_scadas",
+        AsyncMock(return_value=[PROJECT_SCADA.copy()]),
+    )
+
+
+def test_realtime_scada_simulation_uses_current_project_metadata(monkeypatch):
+    _patch_project_scada_metadata(monkeypatch)
+    query_mock = AsyncMock(return_value=[{"time": START_TIME, "value": 26.5}])
+    monkeypatch.setattr(
+        composite_queries.RealtimeRepository,
+        "get_node_field_by_time_range",
+        query_mock,
+        raising=False,
+    )
+
+    result = asyncio.run(
+        composite_queries.CompositeQueries.get_scada_associated_realtime_simulation_data(
+            object(), object(), [PROJECT_SCADA["id"]], START_TIME, END_TIME
+        )
+    )
+
+    assert result[PROJECT_SCADA["id"]][0]["scada_id"] == PROJECT_SCADA["id"]
+    assert query_mock.await_args.args[1:] == (
+        START_TIME,
+        END_TIME,
+        "J1",
+        "pressure",
+    )
+
+
+def test_scheme_scada_simulation_uses_current_project_metadata(monkeypatch):
+    _patch_project_scada_metadata(monkeypatch)
+    query_mock = AsyncMock(return_value=[{"time": START_TIME, "value": 26.5}])
+    monkeypatch.setattr(
+        composite_queries.SchemeRepository,
+        "get_node_field_by_scheme_and_time_range",
+        query_mock,
+        raising=False,
+    )
+
+    result = asyncio.run(
+        composite_queries.CompositeQueries.get_scada_associated_scheme_simulation_data(
+            object(),
+            object(),
+            [PROJECT_SCADA["id"]],
+            START_TIME,
+            END_TIME,
+            "baseline",
+            "scheme-1",
+        )
+    )
+
+    assert result[PROJECT_SCADA["id"]][0]["scada_id"] == PROJECT_SCADA["id"]
+    assert query_mock.await_args.args[5:] == ("J1", "pressure")
+
+
+def test_element_scada_query_uses_current_project_metadata(monkeypatch):
+    _patch_project_scada_metadata(monkeypatch)
+    query_mock = AsyncMock(
+        return_value={PROJECT_SCADA["id"]: [{"time": START_TIME, "value": 26.5}]}
+    )
+    monkeypatch.setattr(
+        composite_queries.ScadaRepository,
+        "get_scada_field_by_id_time_range",
+        query_mock,
+    )
+
+    result = asyncio.run(
+        composite_queries.CompositeQueries.get_element_associated_scada_data(
+            object(), object(), "J1", START_TIME, END_TIME
+        )
+    )
+
+    assert result == {"J1": [{"time": START_TIME, "value": 26.5}]}
+
+
+def test_scada_info_endpoint_uses_current_project_connection(monkeypatch):
+    monkeypatch.setattr(
+        project_data.ScadaInfoRepository,
+        "get_scadas",
+        AsyncMock(return_value=[PROJECT_SCADA.copy()]),
+    )
+
+    result = asyncio.run(project_data.get_scada_info_with_connection(object()))
+
+    assert result == {"success": True, "data": [PROJECT_SCADA], "count": 1}