add load_profile stuff

load_profile_utilities.py

@@ -84,36 +84,195 @@ def plot_profiles(individual_profiles: np.ndarray, aggregated_profile: list, noi
 
 
 
+# if __name__ == "__main__":
+#     # Exemple d'utilisation
+#     hours = range(24)
+#     office_profile = [0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.3, 0.5, 0.7, 0.8, 0.8, 0.8, 0.3, 0.3, 0.8, 0.8, 0.8, 0.7, 0.6, 0.2, 0.1, 0.1, 0.1]
+#     # Ajouter du bruit au profil agrégé
+#     noisy_office = add_noise_to_profile(office_profile, noise_std=0.05)
+#     # Générer les profils individuels pour 5 ordinateurs
+#     N = 10
+#     individual_profiles = generate_computer_profiles(N, noisy_office)
+#     # Afficher les profils
+#     #plot_profiles(individual_profiles, office_profile, noisy_office)
+
+#     HVAC_profile = [0.1, 0.1, 0.1, 0.5, 1, 1, 0.8, 0.7, 0.5, 0.3, 0.3, 0.3, 0.4, 0.7, 0.7, 0.5, 0.3, 0.2, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1]
+#     noisy_HVAC = add_noise_to_profile(HVAC_profile, noise_std=0.05)
+
+#     EV_profile = [0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.3, 0.9, 0.85, 0.8, 0.75, 0.7, 0.3, 0.25, 0.75, 0.7, 0.65, 0.6, 0.5, 0.2, 0.1, 0.1, 0.1]
+#     noisy_EV = add_noise_to_profile(EV_profile, noise_std=0.05)
+
+#     industrial_load_24h = [0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9]
+#     noisy_industrial_load_24h = add_noise_to_profile(industrial_load_24h, noise_std=0.05)
+
+#     industrial_load_12h = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0, 0, 0, 0]
+#     noisy_industrial_load_12h = add_noise_to_profile(industrial_load_12h, noise_std=0.05)
+
+#     plt.figure()
+#     #plt.plot(hours,EV_profile,'g')
+#     #plt.plot(hours,noisy_HVAC,'r')
+#     #plt.plot(hours,noisy_office,'b')
+#     #plt.plot(hours,noisy_HVAC,'k')
+#     #plt.plot(hours,noisy_industrial_load_24h,'y')
+#     plt.plot(hours,noisy_industrial_load_12h,'c')
+#     plt.grid()
+#     plt.show()
+
+
+
+import numpy as np
+from datetime import datetime, timedelta, date as date_obj
+
+def add_noise_to_profile(profile: list, noise_std: float = 0.05) -> np.ndarray:
+    """
+    Ajoute du bruit gaussien à un profil de consommation.
+    """
+    profile = np.array(profile)
+    noise = np.random.normal(0, noise_std, profile.shape)
+    noisy_profile = profile + noise
+    return np.clip(noisy_profile, 0, None)
+
+def get_season(date: date_obj) -> str:
+    """
+    Détermine la saison en fonction du mois de la date.
+    """
+    month = date.month
+    if 3 <= month <= 5:
+        return 'spring'
+    elif 6 <= month <= 8:
+        return 'summer'
+    elif 9 <= month <= 11:
+        return 'autumn'
+    else:
+        return 'winter'
+
+def get_day_type(date: date_obj, holidays: set) -> str:
+    """
+    Détermine si la date est un jour de semaine, un week-end ou des vacances.
+    """
+    if date in holidays:
+        return 'vacation'
+    elif date.weekday() >= 5:
+        return 'weekend'
+    else:
+        return 'workday'
+
+def interpolate_profile(base_profile: list, time_step_min: int) -> np.ndarray:
+    """
+    Interpole un profil de 24 heures à une résolution temporelle plus fine.
+    """
+    hours = np.arange(24)
+    step_per_hour = 60 // time_step_min
+    x_new = np.linspace(0, 23, 24 * step_per_hour)
+    return np.interp(x_new, hours, base_profile)
+
+def generate_load_profile(
+    num_days: int,
+    time_step_min: int = 60,
+    base_profiles: dict = None,
+    holidays: list = None,
+    noise_std: float = 0.05
+) -> tuple:
+    """
+    Génère un profil de charge électrique sur une période donnée.
+
+    Args:
+        num_days (int): Nombre de jours pour lesquels générer le profil.
+        time_step_min (int): Pas de temps en minutes (par défaut 60 minutes).
+        base_profiles (dict): Profils de base pour chaque saison et type de jour.
+        holidays (list): Liste des dates de vacances.
+        noise_std (float): Écart-type du bruit gaussien.
+
+    Returns:
+        tuple: (timestamps, load_values, timestamp_indices) où :
+            - timestamps est un tableau de datetime,
+            - load_values est un tableau de valeurs de charge,
+            - timestamp_indices est un tableau d'indices séquentiels.
+    """
+    # Date de début fixée au 01/01/2018
+    start_date = date_obj(2018, 1, 1)
+    end_date = start_date + timedelta(days=num_days - 1)
+
+    dates = [start_date + timedelta(days=i) for i in range(num_days)]
+
+    timestamps = []
+    load_values = []
+    timestamp_indices = []
+    current_index = 0
+
+    if base_profiles is None:
+        # Profils de base par défaut (à personnaliser selon vos besoins)
+        office_profile = [0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.3, 0.5, 0.7, 0.8, 0.8, 
+                          0.8, 0.3, 0.3, 0.8, 0.8, 0.8, 0.7, 0.6, 0.2, 0.1, 0.1, 0.1]
+        weekend_profile = [0.05] * 24
+        vacation_profile = [0.08] * 24
+
+        base_profiles = {
+            'winter': {'workday': office_profile, 'weekend': weekend_profile, 'vacation': vacation_profile},
+            'spring': {'workday': office_profile, 'weekend': weekend_profile, 'vacation': vacation_profile},
+            'summer': {'workday': office_profile, 'weekend': weekend_profile, 'vacation': vacation_profile},
+            'autumn': {'workday': office_profile, 'weekend': weekend_profile, 'vacation': vacation_profile}
+        }
+
+    holidays = set(holidays) if holidays else set()
+
+    for date in dates:
+        season = get_season(date)
+        day_type = get_day_type(date, holidays)
+
+        try:
+            base_profile = base_profiles[season][day_type]
+        except KeyError as e:
+            raise ValueError(f"Profil manquant pour {season} et {day_type}") from e
+
+        # Générer les timestamps pour la journée
+        steps_per_day = 24 * 60 // time_step_min
+        daily_timestamps = [
+            datetime(date.year, date.month, date.day, (i * time_step_min) // 60, (i * time_step_min) % 60)
+            for i in range(steps_per_day)
+        ]
+
+        # Générer les valeurs de charge
+        if time_step_min == 60:
+            daily_load = add_noise_to_profile(base_profile, noise_std)
+        else:
+            interpolated = interpolate_profile(base_profile, time_step_min)
+            noise = np.random.normal(0, noise_std, interpolated.shape)
+            daily_load = np.clip(interpolated + noise, 0, None)
+
+        # Ajouter les timestamps, les valeurs de charge et les indices
+        timestamps.extend(daily_timestamps)
+        load_values.extend(daily_load)
+        timestamp_indices.extend(range(current_index, current_index + steps_per_day))
+        current_index += steps_per_day
+
+    return np.array(timestamps), np.array(load_values), np.array(timestamp_indices)
+
+# Exemple d'utilisation
 if __name__ == "__main__":
-    # Exemple d'utilisation
-    hours = range(24)
-    office_profile = [0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.3, 0.5, 0.7, 0.8, 0.8, 0.8, 0.3, 0.3, 0.8, 0.8, 0.8, 0.7, 0.6, 0.2, 0.1, 0.1, 0.1]
-    # Ajouter du bruit au profil agrégé
-    noisy_office = add_noise_to_profile(office_profile, noise_std=0.05)
-    # Générer les profils individuels pour 5 ordinateurs
-    N = 10
-    individual_profiles = generate_computer_profiles(N, noisy_office)
-    # Afficher les profils
-    #plot_profiles(individual_profiles, office_profile, noisy_office)
-
-    HVAC_profile = [0.1, 0.1, 0.1, 0.5, 1, 1, 0.8, 0.7, 0.5, 0.3, 0.3, 0.3, 0.4, 0.7, 0.7, 0.5, 0.3, 0.2, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1]
-    noisy_HVAC = add_noise_to_profile(HVAC_profile, noise_std=0.05)
-
-    EV_profile = [0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.3, 0.9, 0.85, 0.8, 0.75, 0.7, 0.3, 0.25, 0.75, 0.7, 0.65, 0.6, 0.5, 0.2, 0.1, 0.1, 0.1]
-    noisy_EV = add_noise_to_profile(EV_profile, noise_std=0.05)
-
-    industrial_load_24h = [0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9]
-    noisy_industrial_load_24h = add_noise_to_profile(industrial_load_24h, noise_std=0.05)
-
-    industrial_load_12h = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0, 0, 0, 0]
-    noisy_industrial_load_12h = add_noise_to_profile(industrial_load_12h, noise_std=0.05)
-
-    plt.figure()
-    #plt.plot(hours,EV_profile,'g')
-    #plt.plot(hours,noisy_HVAC,'r')
-    #plt.plot(hours,noisy_office,'b')
-    #plt.plot(hours,noisy_HVAC,'k')
-    #plt.plot(hours,noisy_industrial_load_24h,'y')
-    plt.plot(hours,noisy_industrial_load_12h,'c')
-    plt.grid()
+    # Exemple de vacances
+    holidays = [date_obj(2018, 12, 25)]  # Noël 2018
+
+    # Générer un profil pour 10 jours avec un pas de temps de 30 minutes
+    timestamps, load, indices = generate_load_profile(
+        num_days=1,
+        time_step_min=5,
+        holidays=holidays,
+        noise_std=0.05
+    )
+
+    # Affichage des 10 premiers éléments pour vérification
+    print("Timestamps:", timestamps[:10])
+    print("Load Values:", load[:10])
+    print("Timestamp Indices:", indices[:10])
+
+    # Affichage du profil généré
+    import matplotlib.pyplot as plt
+    plt.figure(figsize=(12, 6))
+    plt.plot(indices, load, label='Profil de charge')
+    plt.xlabel('Temps')
+    plt.ylabel('Consommation normalisée')
+    plt.title('Profil de charge électrique sur 10 jours (pas de 30 minutes)')
+    plt.grid(True)
+    plt.legend()
     plt.show()

main.py

@@ -6,8 +6,8 @@
 from worst_case_utilities import perform_comprehensive_sizing,validate_network_performance
 from tqdm import tqdm
 
-path = 'grid_data_input_file_building_demo.xlsx'
-# path = 'grid_data_input_file_WIP_v1.xlsx'
+#path = 'grid_data_input_file_building_demo.xlsx'
+path = 'grid_data_input_file_WIP_v1.xlsx'
 path_cable_catalogue = "cable_catalogue.xlsx"
 path_converter_catalogue = "Converters_Library.xlsx"