Development of an AI system for predictive maintenance of power grids
Power equipment—transformers, cables, switches—fails with early warning signs of degradation weeks and months before an accident. Shifting from preventive to predictive maintenance reduces costs by 20-30% and reduces unplanned outages.
Diagnostics of power transformers
DGA (Dissolved Gas Analysis) is the main indicator:
import numpy as np
# Ключи диагностики: растворённые газы в масле трансформатора
dga_thresholds = {
'hydrogen_H2': {'warning': 150, 'alarm': 500}, # ppm
'methane_CH4': {'warning': 75, 'alarm': 200},
'ethylene_C2H4': {'warning': 60, 'alarm': 150},
'ethane_C2H6': {'warning': 100, 'alarm': 200},
'acetylene_C2H2': {'warning': 2, 'alarm': 30}, # ацетилен = дуга!
'carbon_monoxide_CO': {'warning': 700, 'alarm': 1500},
'carbon_dioxide_CO2': {'warning': 10000, 'alarm': 15000}
}
def diagnose_transformer_dga(gas_ppm: dict) -> dict:
"""
Треугольник Дюваля + Rogers ratios для классификации дефекта.
"""
# Проверка порогов
alarm_gases = []
for gas, value in gas_ppm.items():
if gas in dga_thresholds:
if value > dga_thresholds[gas]['alarm']:
alarm_gases.append({'gas': gas, 'value': value, 'level': 'alarm'})
elif value > dga_thresholds[gas]['warning']:
alarm_gases.append({'gas': gas, 'value': value, 'level': 'warning'})
# Rogers ratios (IEEE C57.104)
ch4_h2 = gas_ppm.get('methane_CH4', 0) / (gas_ppm.get('hydrogen_H2', 1) + 1e-9)
c2h4_c2h6 = gas_ppm.get('ethylene_C2H4', 0) / (gas_ppm.get('ethane_C2H6', 1) + 1e-9)
c2h2_c2h4 = gas_ppm.get('acetylene_C2H2', 0) / (gas_ppm.get('ethylene_C2H4', 1) + 1e-9)
# Диагностика дефекта
fault_type = 'normal'
if gas_ppm.get('acetylene_C2H2', 0) > 5:
fault_type = 'arc_discharge' # электрическая дуга — КРИТИЧНО
elif c2h4_c2h6 > 1.0 and ch4_h2 > 0.1:
fault_type = 'thermal_fault_high' # перегрев > 700°C
elif ch4_h2 > 0.1 and c2h4_c2h6 < 1.0:
fault_type = 'thermal_fault_medium' # перегрев 300-700°C
elif gas_ppm.get('hydrogen_H2', 0) > 200 and c2h4_c2h6 < 0.1:
fault_type = 'partial_discharge' # частичные разряды
urgency = {
'arc_discharge': 'immediate_shutdown',
'thermal_fault_high': 'urgent_inspection',
'thermal_fault_medium': 'schedule_maintenance',
'partial_discharge': 'enhanced_monitoring',
'normal': 'routine'
}
return {
'fault_type': fault_type,
'urgency': urgency[fault_type],
'alarm_gases': alarm_gases,
'rogers_ratios': {
'CH4/H2': round(ch4_h2, 3),
'C2H4/C2H6': round(c2h4_c2h6, 3),
'C2H2/C2H4': round(c2h2_c2h4, 3)
}
}
Gas Trend - Degradation Forecast:
from sklearn.linear_model import LinearRegression
import pandas as pd
def forecast_gas_trend(dga_history: pd.DataFrame, gas: str,
forecast_days: int = 30) -> dict:
"""
Экспоненциальный рост газов = ускорение дефекта.
Линейный тренд на log(concentration) для обнаружения ускорения.
"""
data = dga_history[['days_ago', gas]].dropna()
data = data[data[gas] > 0]
if len(data) < 3:
return {'status': 'insufficient_data'}
X = data['days_ago'].values.reshape(-1, 1)
y = np.log(data[gas].values)
model = LinearRegression()
model.fit(X, y)
# Прогноз
future_x = np.array([[-forecast_days]]) # отрицательный = будущее
predicted_log = model.predict(future_x)[0]
predicted_concentration = np.exp(predicted_log)
doubling_time = np.log(2) / abs(model.coef_[0]) if model.coef_[0] > 0 else None
return {
'current': data[gas].iloc[-1],
f'forecast_{forecast_days}d': round(predicted_concentration, 1),
'growth_rate_per_day': model.coef_[0],
'doubling_time_days': round(doubling_time, 0) if doubling_time else None,
'threshold_breach_days': estimate_days_to_threshold(
data[gas].iloc[-1], predicted_concentration,
dga_thresholds.get(gas, {}).get('alarm', float('inf')),
forecast_days
)
}
Monitoring of cable lines
Partial discharges in the cable:
def analyze_partial_discharge(pd_measurements: dict) -> dict:
"""
PD-мониторинг высоковольтных кабелей.
pC (пикокулоны) — величина разряда.
PDIV (Partial Discharge Inception Voltage) — напряжение начала ПР.
"""
max_pd_pc = pd_measurements.get('max_pd_magnitude_pc', 0)
pd_rate_per_minute = pd_measurements.get('pd_pulse_rate', 0)
pd_location_m = pd_measurements.get('pd_location_tdr_m', None) # TDR локация
# Классификация по IEC 60270
if max_pd_pc > 1000:
severity = 'critical'
action = 'immediate_cable_replacement'
elif max_pd_pc > 300:
severity = 'major'
action = 'schedule_replacement_3months'
elif max_pd_pc > 100:
severity = 'minor'
action = 'enhanced_monitoring'
else:
severity = 'normal'
action = 'routine_monitoring'
return {
'max_pd_pc': max_pd_pc,
'pd_rate': pd_rate_per_minute,
'pd_location_m': pd_location_m,
'severity': severity,
'action': action
}
Predictive model for overhead lines
Wire sag prediction:
def calculate_sag_risk(weather_data: dict, line_parameters: dict) -> dict:
"""
Нагрев провода током + температура → провисание → опасное приближение к земле.
Допустимый нагрев AAAC провода: до 90°C (аварийный режим).
"""
I_amps = line_parameters['current_a']
R_ohm_per_km = line_parameters['resistance_ohm_per_km']
length_km = line_parameters['length_km']
ambient_temp = weather_data['temperature_c']
wind_speed = weather_data['wind_speed_ms']
solar_radiation = weather_data.get('solar_irradiance_wm2', 0)
# Упрощённая тепловая модель (IEEE 738)
joule_heating = I_amps**2 * R_ohm_per_km / 1000 # Вт/м
convective_cooling = (0.7 + 0.5 * wind_speed) * (80 - ambient_temp) # Вт/м (упрощение)
solar_heating = solar_radiation * 0.015 # Вт/м
conductor_temp = ambient_temp + (joule_heating + solar_heating) / (convective_cooling + 1)
# Провисание растёт с температурой
base_sag_m = line_parameters.get('design_sag_m', 5.0)
thermal_expansion = 0.023e-3 * (conductor_temp - 20) * length_km * 1000
actual_sag_m = base_sag_m + thermal_expansion * 0.5
clearance_violation = actual_sag_m > line_parameters.get('max_sag_m', 7.0)
return {
'conductor_temp_c': round(conductor_temp, 1),
'actual_sag_m': round(actual_sag_m, 2),
'clearance_violation': clearance_violation,
'thermal_limit_pct': min(100, conductor_temp / 90 * 100)
}
Maintenance planning
Prioritization of objects: Risk matrix: failure probability × consequences (kWh of lost output × fines × consumers). Facilities are ranked by expected losses from an unplanned failure. Integration with SAP PM / Maximo for work order creation.
Deadlines: DGA monitoring + basic alerting + transformer dashboard — 4-5 weeks. DGA trends, cable PD monitoring, power transmission line thermal model, maintenance planning optimization, SAP PM integration — 3-4 months.