|
|
@@ -1,121 +1,124 @@
|
|
|
import pandas as pd
|
|
|
import os
|
|
|
-import plotly.graph_objects as go
|
|
|
from algorithmContract.confBusiness import *
|
|
|
from algorithmContract.contract import Contract
|
|
|
from behavior.analystNotFilter import AnalystNotFilter
|
|
|
-from plotly.subplots import make_subplots
|
|
|
|
|
|
class FaultAnalyst(AnalystNotFilter):
|
|
|
"""
|
|
|
风电机组故障分析
|
|
|
+ 新逻辑：基于首发故障结束时间划分窗口， 以首发故障的结束时间为界，后续故障的开始时间如果 ≤ 这个结束时间，则视为连带故障；否则作为新的首发故障。
|
|
|
+
|
|
|
+ 风场范围的故障分析
|
|
|
+ 统计风场范围内 首发故障中各故障发生的次数和累计时长
|
|
|
"""
|
|
|
|
|
|
def typeAnalyst(self):
|
|
|
return "fault"
|
|
|
-
|
|
|
- def selectColumns(self):
|
|
|
- # 这里的字段必须与数据库 _fault 表中的列名对应
|
|
|
- return [Field_DeviceName, Field_FaultTime, Field_FaultDetail]
|
|
|
|
|
|
- # 强制重写获取数据源类型的方法，防止因配置错误去查 _minute 表，强制代码去查 _fault 表
|
|
|
+ def selectColumns(self):
|
|
|
+ # 直接返回数据库表中的真实列名
|
|
|
+ return ["wind_turbine_name", "begin_time", "end_time", "fault_detail"]
|
|
|
+ #风机原始名称、故障开始时间、故障结束时间、故障描述
|
|
|
def getTimeGranularitys(self, conf: Contract):
|
|
|
- return ["fault"]
|
|
|
+ return ["fault"]
|
|
|
|
|
|
def turbinesAnalysis(self, outputAnalysisDir, conf: Contract, turbineCodes):
|
|
|
- # 1. 抓取数据
|
|
|
dictionary = self.processTurbineData(turbineCodes, conf, self.selectColumns())
|
|
|
-
|
|
|
- # 2. 直接获取 "fault" key 的数据, 绕过父类的 userDataFrame 方法，因为它会读取错误的配置文件 ('minute')
|
|
|
dataFrameMerge = dictionary.get("fault", pd.DataFrame())
|
|
|
-
|
|
|
- # 3. 增加空数据保护
|
|
|
if dataFrameMerge.empty:
|
|
|
- # print("Warning: No fault data found for the selected turbines.")
|
|
|
return pd.DataFrame()
|
|
|
-
|
|
|
return self.get_result(dataFrameMerge, outputAnalysisDir, conf)
|
|
|
|
|
|
def get_result(self, dataFrame: pd.DataFrame, outputAnalysisDir: str, conf: Contract):
|
|
|
- # 双重保险：如果数据为空直接返回
|
|
|
if dataFrame.empty:
|
|
|
return pd.DataFrame()
|
|
|
|
|
|
- #---------------整个风场维度统计故障时长与次数---------------------------
|
|
|
- # 统计各种类型故障出现的次数
|
|
|
- if Field_FaultDetail in dataFrame.columns:
|
|
|
- fault_detail_count = dataFrame[Field_FaultDetail].value_counts().reset_index()
|
|
|
- fault_detail_count.columns = [Field_FaultDetail, 'count']
|
|
|
+ df = dataFrame.copy()
|
|
|
+
|
|
|
+ # 转换时间列
|
|
|
+ df["begin_time"] = pd.to_datetime(df["begin_time"])
|
|
|
+ df["end_time"] = pd.to_datetime(df["end_time"])
|
|
|
+ df = df.dropna(subset=["begin_time", "end_time"])
|
|
|
|
|
|
- # 统计每个 fault_detail 的时长加和
|
|
|
- fault_time_sum = dataFrame.groupby(Field_FaultDetail)[Field_FaultTime].sum().reset_index()
|
|
|
- fault_time_sum.columns = [Field_FaultDetail, 'fault_time_sum']
|
|
|
+ if df.empty:
|
|
|
+ return pd.DataFrame()
|
|
|
|
|
|
- # 合并两个 DataFrame
|
|
|
- fault_summary = pd.merge(fault_detail_count, fault_time_sum, on=Field_FaultDetail, how='inner')
|
|
|
- fault_summary_sorted = fault_summary.sort_values(by='fault_time_sum', ascending=False)
|
|
|
+ # 存储中间结果
|
|
|
+ turbine_events = []
|
|
|
+ fault_events = []
|
|
|
+
|
|
|
+ # 按风机分组处理
|
|
|
+ for turbine, group in df.groupby("wind_turbine_name"):
|
|
|
+ group = group.sort_values("begin_time").reset_index(drop=True)
|
|
|
+ i = 0
|
|
|
+ n = len(group)
|
|
|
+ while i < n:
|
|
|
+ primary = group.iloc[i]
|
|
|
+ primary_start = primary["begin_time"]
|
|
|
+ primary_end = primary["end_time"]
|
|
|
+ duration_sec = (primary_end - primary_start).total_seconds()
|
|
|
+
|
|
|
+ turbine_events.append({
|
|
|
+ "turbine": turbine,
|
|
|
+ "duration_sec": duration_sec
|
|
|
+ })
|
|
|
+ fault_events.append({
|
|
|
+ "fault_detail": primary["fault_detail"],
|
|
|
+ "duration_sec": duration_sec
|
|
|
+ })
|
|
|
+
|
|
|
+ # 跳过连带故障
|
|
|
+ j = i + 1
|
|
|
+ while j < n and group.iloc[j]["begin_time"] <= primary_end:
|
|
|
+ j += 1
|
|
|
+ i = j
|
|
|
+
|
|
|
+ # 聚合风机维度
|
|
|
+ if turbine_events:
|
|
|
+ turbine_df = pd.DataFrame(turbine_events)
|
|
|
+ turbine_summary = turbine_df.groupby("turbine").agg(
|
|
|
+ count=("duration_sec", "size"),
|
|
|
+ fault_time=("duration_sec", "sum")
|
|
|
+ ).reset_index()
|
|
|
+ turbine_summary = turbine_summary.rename(columns={"turbine": "wind_turbine_name"})
|
|
|
+ turbine_file = os.path.join(outputAnalysisDir, f"turbine_fault_result{CSVSuffix}")
|
|
|
+ turbine_summary.to_csv(turbine_file, index=False, encoding='utf-8-sig')
|
|
|
else:
|
|
|
- # 防御性代码：如果缺列
|
|
|
- fault_summary_sorted = pd.DataFrame()
|
|
|
-
|
|
|
- # -------------按风机分组统计故障情况------------------------------------------
|
|
|
- # 确保有设备名称列
|
|
|
- if Field_DeviceName not in dataFrame.columns:
|
|
|
- # 有时 Field_DeviceName 没取到，尝试用 DeviceCode 或其他
|
|
|
- groupby_col = dataFrame.columns[0]
|
|
|
+ turbine_summary = pd.DataFrame()
|
|
|
+
|
|
|
+ # 聚合故障类型维度
|
|
|
+ if fault_events:
|
|
|
+ fault_df = pd.DataFrame(fault_events)
|
|
|
+ fault_summary = fault_df.groupby("fault_detail").agg(
|
|
|
+ count=("duration_sec", "size"),
|
|
|
+ fault_time_sum=("duration_sec", "sum")
|
|
|
+ ).reset_index()
|
|
|
+ fault_file = os.path.join(outputAnalysisDir, f"total_fault_result{CSVSuffix}")
|
|
|
+ fault_summary.to_csv(fault_file, index=False, encoding='utf-8-sig')
|
|
|
else:
|
|
|
- groupby_col = Field_DeviceName
|
|
|
-
|
|
|
- grouped = dataFrame.groupby(groupby_col)
|
|
|
- results= []
|
|
|
+ fault_summary = pd.DataFrame()
|
|
|
|
|
|
- for name, group in grouped:
|
|
|
- turbine_fault_summary = pd.DataFrame({
|
|
|
- Field_DeviceName: [name],
|
|
|
- 'count': [len(group)],
|
|
|
- 'fault_time': [group[Field_FaultTime].sum()]
|
|
|
- })
|
|
|
- results.append(turbine_fault_summary)
|
|
|
-
|
|
|
- # 合并所有风机的故障统计结果
|
|
|
- if results:
|
|
|
- turbine_fault_summary = pd.concat(results, ignore_index=True)
|
|
|
- turbine_fault_sorted = turbine_fault_summary.sort_values(by='fault_time', ascending=False)
|
|
|
- # 故障类型前十名
|
|
|
- # draw_results=turbine_fault_sorted.head(10) # 暂时没用到
|
|
|
- else:
|
|
|
- turbine_fault_sorted = pd.DataFrame()
|
|
|
-
|
|
|
- # 保存结果
|
|
|
+ # 返回结果
|
|
|
result_rows = []
|
|
|
-
|
|
|
- if not turbine_fault_sorted.empty:
|
|
|
- filePathOfturbinefault = os.path.join(outputAnalysisDir, f"turbine_fault_result{CSVSuffix}")
|
|
|
- turbine_fault_sorted.to_csv(filePathOfturbinefault, index=False,encoding='utf-8-sig')
|
|
|
-
|
|
|
+ if not turbine_summary.empty:
|
|
|
result_rows.append({
|
|
|
Field_Return_TypeAnalyst: self.typeAnalyst(),
|
|
|
Field_PowerFarmCode: conf.dataContract.dataFilter.powerFarmID,
|
|
|
Field_Return_BatchCode: conf.dataContract.dataFilter.dataBatchNum,
|
|
|
Field_CodeOfTurbine: "total",
|
|
|
- Field_MillTypeCode:"turbine_fault_result",
|
|
|
- Field_Return_FilePath: filePathOfturbinefault,
|
|
|
+ Field_MillTypeCode: "turbine_fault_result",
|
|
|
+ Field_Return_FilePath: turbine_file,
|
|
|
Field_Return_IsSaveDatabase: True
|
|
|
})
|
|
|
-
|
|
|
- if not fault_summary_sorted.empty:
|
|
|
- filePathOftotalfault = os.path.join(outputAnalysisDir, f"total_fault_result{CSVSuffix}")
|
|
|
- fault_summary_sorted.to_csv(filePathOftotalfault, index=False,encoding='utf-8-sig')
|
|
|
-
|
|
|
+ if not fault_summary.empty:
|
|
|
result_rows.append({
|
|
|
Field_Return_TypeAnalyst: self.typeAnalyst(),
|
|
|
Field_PowerFarmCode: conf.dataContract.dataFilter.powerFarmID,
|
|
|
Field_Return_BatchCode: conf.dataContract.dataFilter.dataBatchNum,
|
|
|
Field_CodeOfTurbine: "total",
|
|
|
- Field_MillTypeCode:"total_fault_result",
|
|
|
- Field_Return_FilePath: filePathOftotalfault,
|
|
|
+ Field_MillTypeCode: "total_fault_result",
|
|
|
+ Field_Return_FilePath: fault_file,
|
|
|
Field_Return_IsSaveDatabase: True
|
|
|
})
|
|
|
-
|
|
|
- result_df = pd.DataFrame(result_rows)
|
|
|
- return result_df
|
|
|
+ return pd.DataFrame(result_rows)
|
wangjiaojiao