基于KNN的旋转机械故障诊断（Python）

最新推荐文章于 2024-07-09 23:19:55 发布

哥廷根数学学派

最新推荐文章于 2024-07-09 23:19:55 发布

阅读量286

点赞数 8

分类专栏：信号处理机器学习故障诊断文章标签： python 开发语言机器学习信号处理算法

本文链接：https://blog.csdn.net/weixin_39402231/article/details/139953212

版权

信号处理同时被 3 个专栏收录

301 篇文章 23 订阅

订阅专栏

机器学习

157 篇文章 1 订阅

订阅专栏

故障诊断

119 篇文章 46 订阅

订阅专栏

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.model_selection import train_test_split
from sklearn.preprocessing import MinMaxScaler
from sklearn.neighbors import KNeighborsClassifier
from sklearn.metrics import classification_report
from sklearn.metrics import confusion_matrix
import sklearn.externals
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

scaler = MinMaxScaler()
# Escalamos los predictores:
X_train = scaler.fit_transform(X_train)
X_test = scaler.transform(X_test)

k_values = range(1, 30)
scores_train = []
scores_test = []
for k in k_values:
    knn = KNeighborsClassifier(n_neighbors = k)
    knn.fit(X_train, y_train)
    scores_train.append(knn.score(X_train, y_train))
    scores_test.append(knn.score(X_test, y_test))
    print('----- k: ' + str(k) + '----- Accuracy test: ' + str(scores_test[k - 1]) + '-----')
plt.figure()
plt.xlabel('k')
plt.ylabel('Accuracy')
plt.plot(k_values, scores_train, label = 'Train')
plt.plot(k_values, scores_test, label = 'Test')
plt.legend()

----- k: 1----- Accuracy test: 0.9777777777777777-----
----- k: 2----- Accuracy test: 0.9666666666666667-----
----- k: 3----- Accuracy test: 0.9777777777777777-----
----- k: 4----- Accuracy test: 0.9777777777777777-----
----- k: 5----- Accuracy test: 0.9777777777777777-----
----- k: 6----- Accuracy test: 0.9777777777777777-----
----- k: 7----- Accuracy test: 0.9777777777777777-----
----- k: 8----- Accuracy test: 0.9777777777777777-----
----- k: 9----- Accuracy test: 0.9666666666666667-----
----- k: 10----- Accuracy test: 0.9666666666666667-----
----- k: 11----- Accuracy test: 0.9444444444444444-----
----- k: 12----- Accuracy test: 0.9444444444444444-----
----- k: 13----- Accuracy test: 0.9444444444444444-----
----- k: 14----- Accuracy test: 0.9555555555555556-----
----- k: 15----- Accuracy test: 0.9444444444444444-----
----- k: 16----- Accuracy test: 0.9555555555555556-----
----- k: 17----- Accuracy test: 0.9555555555555556-----
----- k: 18----- Accuracy test: 0.9555555555555556-----
----- k: 19----- Accuracy test: 0.9444444444444444-----
----- k: 20----- Accuracy test: 0.9555555555555556-----
----- k: 21----- Accuracy test: 0.9555555555555556-----
----- k: 22----- Accuracy test: 0.9555555555555556-----
----- k: 23----- Accuracy test: 0.9555555555555556-----
----- k: 24----- Accuracy test: 0.9555555555555556-----
----- k: 25----- Accuracy test: 0.9555555555555556-----
----- k: 26----- Accuracy test: 0.9555555555555556-----
----- k: 27----- Accuracy test: 0.9555555555555556-----
----- k: 28----- Accuracy test: 0.9555555555555556-----
----- k: 29----- Accuracy test: 0.9555555555555556-----

n_neighbors = 5
classifier = KNeighborsClassifier(n_neighbors)
classifier.fit(X_train, y_train)
print('Accuracy en el conjunto de train: {:.3f}'
     .format(classifier.score(X_train, y_train)))
print('Accuracy en el conjunto de test: {:.3f}'
     .format(classifier.score(X_test, y_test)))

Accuracy en el conjunto de train: 0.981
Accuracy en el conjunto de test: 0.978

target_names = ['Inner', 'Outer', 'Healthy']
pred = classifier.predict(X_test)
print(confusion_matrix(y_test, pred))
print(classification_report(y_test, pred, target_names = target_names))

[[29  1  0]
 [ 0 30  0]
 [ 0  1 29]]
              precision    recall  f1-score   support

       Inner       1.00      0.97      0.98        30
       Outer       0.94      1.00      0.97        30
     Healthy       1.00      0.97      0.98        30

    accuracy                           0.98        90
   macro avg       0.98      0.98      0.98        90
weighted avg       0.98      0.98      0.98        90

pred_train = classifier.predict(X_train)
print(confusion_matrix(y_train, pred_train))
print(classification_report(y_train, pred_train, target_names = target_names))

[[90  0  0]
 [ 0 86  4]
 [ 0  1 89]]
              precision    recall  f1-score   support

       Inner       1.00      1.00      1.00        90
       Outer       0.99      0.96      0.97        90
     Healthy       0.96      0.99      0.97        90

    accuracy                           0.98       270
   macro avg       0.98      0.98      0.98       270
weighted avg       0.98      0.98      0.98       270

sns.set()
mat = confusion_matrix(y_test, pred)
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=['Inner', 'Outer', 'Healthy'],
            yticklabels=['Inner', 'Outer', 'Healthy'],
            cmap=sns.cubehelix_palette(light=1, as_cmap=True))


plt.xlabel('True label');
plt.ylabel('Predicted label');

sns.set()
mat = confusion_matrix(y_train, classifier.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=['Inner', 'Outer', 'Healthy'],
            yticklabels=['Inner', 'Outer', 'Healthy'],
            cmap=sns.cubehelix_palette(light=1, as_cmap=True))


plt.xlabel('True label');
plt.ylabel('Predicted label');

joblib.dump(classifier, 'knn_traintest_trained.pkl')

['knn_traintest_trained.pkl']

scaler_filename = "scaler_traintest.pkl"
joblib.dump(scaler, scaler_filename) 

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

['scaler_traintest.pkl']

工学博士，担任《Mechanical System and Signal Processing》《中国电机工程学报》《控制与决策》等期刊审稿专家，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。