基于AdaBoost的旋转机械故障诊断（Python）

前置文章：

将一维机械振动信号构造为训练集和测试集（Python）

https://mp.weixin.qq.com/s/DTKjBo6_WAQ7bUPZEdB1TA

旋转机械振动信号特征提取（Python）

https://mp.weixin.qq.com/s/VwvzTzE-pacxqb9rs8hEVw

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
from matplotlib.colors import ListedColormap
import matplotlib.patches as mpatches
from sklearn.ensemble import AdaBoostClassifier
from sklearn.ensemble import RandomForestClassifier
from sklearn.metrics import classification_report
from sklearn.metrics import confusion_matrix
from sklearn import tree
import joblib

df_train = pd.read_csv("statistics_10_train.csv" , sep = ',')
df_test = pd.read_csv("statistics_10_test.csv" , sep = ',')

X_train = df_train[['Kurtosis', 'Impulse factor', 'RMS', 'Margin factor', 'Skewness',
               'Shape factor', 'Peak to peak', 'Crest factor']].values
y_train = df_train['Tipo'].values
X_test = df_test[['Kurtosis', 'Impulse factor', 'RMS', 'Margin factor', 'Skewness',
               'Shape factor', 'Peak to peak', 'Crest factor']].values
y_test = df_test['Tipo'].values

n_estimators = range(1,150)
scores_train = []
scores_test = []
for i in n_estimators:
    abModel = AdaBoostClassifier(base_estimator = RandomForestClassifier(), n_estimators = i, learning_rate = 0.1)
    abModel.fit(X_train, y_train)
    scores_train.append(abModel.score(X_train, y_train))
    scores_test.append(abModel.score(X_test, y_test))
    if (i % 10 == 0):
        print('----- n estimators: ' + str(i) + '----- Accuracy test: ' + str(scores_test[i - 1]) + '-----')
plt.figure()
plt.xlabel('n_estimators')
plt.ylabel('Accuracy')
plt.plot(n_estimators, scores_train, label = 'Train')
plt.plot(n_estimators, scores_test, label = 'Test')
plt.legend()

n_estimators

range(1, 150)

abModel = AdaBoostClassifier(RandomForestClassifier(n_estimators = 90, min_samples_split = 2, min_samples_leaf  = 1),
                             n_estimators = 125, random_state=0)
abModel.fit(X_train, y_train)
pred_abModel = abModel.predict(X_test)

pred_train_abModel = abModel.predict(X_train)
print(confusion_matrix(y_train, pred_train_abModel))
print(classification_report(y_train, pred_train_abModel))

[[90  0  0]
 [ 0 90  0]
 [ 0  0 90]]
              precision    recall  f1-score   support

       Inner       1.00      1.00      1.00        90
       Outer       1.00      1.00      1.00        90
        Sano       1.00      1.00      1.00        90

    accuracy                           1.00       270
   macro avg       1.00      1.00      1.00       270
weighted avg       1.00      1.00      1.00       270

print(confusion_matrix(y_test, pred_abModel))
print(classification_report(y_test, pred_abModel))

[[28  2  0]
 [ 0 30  0]
 [ 0  3 27]]
              precision    recall  f1-score   support

       Inner       1.00      0.93      0.97        30
       Outer       0.86      1.00      0.92        30
        Sano       1.00      0.90      0.95        30

    accuracy                           0.94        90
   macro avg       0.95      0.94      0.95        90
weighted avg       0.95      0.94      0.95        90

sns.set()
mat = confusion_matrix(y_test, pred_abModel)
fig, ax = plt.subplots(figsize=(7,6))
sns.set(font_scale=1.3)
sns.heatmap(mat.T, square=False, annot=True, fmt='d', cbar=False,
            xticklabels=['Fallo inner race', 'Fallo oute race', 'Sano'],
            yticklabels=['Fallo inner race', 'Fallo oute race', 'Sano'],
            cmap=sns.cubehelix_palette(light=1, as_cmap=True))


plt.xlabel('true label');
plt.ylabel('predicted label');

sns.set()
mat = confusion_matrix(y_train, abModel.predict(X_train))
fig, ax = plt.subplots(figsize=(7,6))
sns.set(font_scale=1.3)
sns.heatmap(mat.T, square=False, annot=True, fmt='d', cbar=False,
            xticklabels=['Fallo inner race', 'Fallo oute race', 'Sano'],
            yticklabels=['Fallo inner race', 'Fallo oute race', 'Sano'],
            cmap=sns.cubehelix_palette(light=1, as_cmap=True))


plt.xlabel('true label');
plt.ylabel('predicted label');

知乎学术咨询：
https://www.zhihu.com/consult/people/792359672131756032?isMe=1

擅长领域：现代信号处理，机器学习，深度学习，数字孪生，时间序列分析，设备缺陷检测、设备异常检测、设备智能故障诊断与健康管理PHM等。