Machine Learning for Drilling Monitoring and Control in Oil Wells: A Case Study for InoVision

InoVision
# Import necessary libraries
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from sklearn.preprocessing import StandardScaler
from sklearn.model_selection import train_test_split
from sklearn.ensemble import RandomForestClassifier
from sklearn.metrics import classification_report, confusion_matrix

# Set seed for reproducibility
np.random.seed(42)

# Step 1: Generate Synthetic Drilling Data
num_samples = 1000
time = np.arange(num_samples)
pressure = np.random.normal(200, 10, num_samples)  # Pressure (psi)
temperature = np.random.normal(80, 5, num_samples)  # Temperature (°C)
rpm = np.random.normal(150, 20, num_samples)  # Rotation speed (RPM)
torque = np.random.normal(5000, 1000, num_samples)  # Torque (Nm)
flow_rate = np.random.normal(500, 50, num_samples)  # Mud flow rate (L/min)

# Create DataFrame
df = pd.DataFrame({
    'Time': time,
    'Pressure': pressure,
    'Temperature': temperature,
    'RPM': rpm,
    'Torque': torque,
    'FlowRate': flow_rate
})

# Add anomaly labels (e.g., pressure surge or stuck pipe)
df['Anomaly'] = 0
df.loc[df['Pressure'] > 220, 'Anomaly'] = 1  # Pressure surge
df.loc[df['Torque'] > 6000, 'Anomaly'] = 1  # Abnormal torque

print(df.head())

# Step 2: Preprocess Data
# Normalize the data
scaler = StandardScaler()
df_scaled = scaler.fit_transform(df[['Pressure', 'Temperature', 'RPM', 'Torque', 'FlowRate']])

# Split data into train and test sets
X = df_scaled
y = df['Anomaly']
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

# Step 3: Build and Train a Machine Learning Model
# Use Random Forest for anomaly detection
model = RandomForestClassifier(n_estimators=100, random_state=42)
model.fit(X_train, y_train)

# Make predictions
y_pred = model.predict(X_test)

# Evaluate the model
print("Classification Report:")
print(classification_report(y_test, y_pred))
print("Confusion Matrix:")
print(confusion_matrix(y_test, y_pred))

# Step 4: Real-Time Monitoring Simulation
# Simulate new data for real-time monitoring
new_data = np.array([[210, 85, 160, 5500, 500]])  # Pressure, Temperature, RPM, Torque, FlowRate
new_data_scaled = scaler.transform(new_data)

# Predict anomalies
prediction = model.predict(new_data_scaled)
print("Anomaly Detected:" if prediction[0] == 1 else "Normal Conditions")

# Step 5: Visualize Results
# Plot pressure data with anomaly detection
plt.figure(figsize=(10, 6))
plt.plot(df['Time'], df['Pressure'], label='Pressure')
plt.scatter(df[df['Anomaly'] == 1]['Time'], df[df['Anomaly'] == 1]['Pressure'], color='red', label='Anomaly')
plt.title('Drilling Pressure Monitoring with Anomaly Detection')
plt.xlabel('Time')
plt.ylabel('Pressure (psi)')
plt.legend()
plt.grid(True)
plt.show()