質問編集履歴
3
誤字の変更
test
CHANGED
|
File without changes
|
test
CHANGED
|
@@ -66,4 +66,4 @@
|
|
|
66
66
|
|
|
67
67
|
plt.yscale('log')
|
|
68
68
|
|
|
69
|
-
plt.show()
|
|
69
|
+
plt.show()'''
|
2
誤字の変更
test
CHANGED
|
File without changes
|
test
CHANGED
|
@@ -8,18 +8,62 @@
|
|
|
8
8
|
|
|
9
9
|
```python
|
|
10
10
|
|
|
11
|
+
#
|
|
12
|
+
|
|
11
|
-
import numpy as np
|
|
13
|
+
import numpy as np #added by author
|
|
14
|
+
|
|
15
|
+
from scipy import signal
|
|
12
16
|
|
|
13
17
|
import matplotlib.pyplot as plt
|
|
14
18
|
|
|
15
|
-
import scipy.stats as sp
|
|
16
19
|
|
|
17
|
-
import math
|
|
18
20
|
|
|
19
|
-
|
|
21
|
+
#Generate a test signal, a 2 Vrms sine wave at 50Hz corrupted by 0.001 V**2/Hz of white noise sampled at 1024 Hz.
|
|
22
|
+
|
|
23
|
+
#テスト信号、1024 Hzでサンプリングされた0.001 V ** 2 / Hzのホワイトノイズで破損した50 Hzの2 Vrmsの正弦波を生成します
|
|
20
24
|
|
|
21
25
|
|
|
22
26
|
|
|
27
|
+
fs = 1024
|
|
28
|
+
|
|
29
|
+
N = 10*fs
|
|
30
|
+
|
|
31
|
+
nperseg = 512
|
|
32
|
+
|
|
33
|
+
amp = 2 * np.sqrt(2)
|
|
34
|
+
|
|
35
|
+
noise_power = 0.001 * fs / 2
|
|
36
|
+
|
|
37
|
+
time = np.arange(N) / float(fs)
|
|
38
|
+
|
|
39
|
+
carrier = amp * np.sin(2*np.pi*50*time)
|
|
40
|
+
|
|
41
|
+
noise = np.random.normal(scale=np.sqrt(noise_power),
|
|
42
|
+
|
|
43
|
+
size=time.shape)
|
|
44
|
+
|
|
45
|
+
x = carrier + noise
|
|
46
|
+
|
|
47
|
+
#Compute and plot the STFT’s magnitude.
|
|
48
|
+
|
|
49
|
+
#STFTの振幅を計算してプロットします
|
|
23
50
|
|
|
24
51
|
|
|
52
|
+
|
|
25
|
-
|
|
53
|
+
f, t, Zxx = signal.stft(x, fs=fs, nperseg=nperseg)
|
|
54
|
+
|
|
55
|
+
plt.figure()
|
|
56
|
+
|
|
57
|
+
plt.pcolormesh(t, f, np.abs(Zxx), vmin=0, vmax=amp)
|
|
58
|
+
|
|
59
|
+
plt.ylim([f[1], f[-1]])
|
|
60
|
+
|
|
61
|
+
plt.title('STFT Magnitude')
|
|
62
|
+
|
|
63
|
+
plt.ylabel('Frequency [Hz]')
|
|
64
|
+
|
|
65
|
+
plt.xlabel('Time [sec]')
|
|
66
|
+
|
|
67
|
+
plt.yscale('log')
|
|
68
|
+
|
|
69
|
+
plt.show()
|
1
誤字の変更
test
CHANGED
|
File without changes
|
test
CHANGED
|
@@ -20,96 +20,6 @@
|
|
|
20
20
|
|
|
21
21
|
|
|
22
22
|
|
|
23
|
-
H, N, eta, frame, shift = 512, 81920, 0.002, 512, 256
|
|
24
|
-
|
|
25
|
-
#タップ数、データ点数、ステップサイズ、フレームサイズ、フレームシフト
|
|
26
|
-
|
|
27
|
-
M, T, C, fs, c = 319, 1500, 30, 16000., 340.
|
|
28
|
-
|
|
29
|
-
#分割数、更新回数、第ｃ周波数成分、サンプリング周波数、音速
|
|
30
23
|
|
|
31
24
|
|
|
32
|
-
|
|
33
|
-
|
|
34
|
-
|
|
35
|
-
s1 = np.loadtxt('../data/a27.txt') #データの読み込み
|
|
36
|
-
|
|
37
|
-
s2 = np.loadtxt('../data/a04.txt')
|
|
38
|
-
|
|
39
|
-
s = np.array([[0 for i in range(N)] for j in range(2)],dtype=float)
|
|
40
|
-
|
|
41
|
-
for i in range(N):
|
|
42
|
-
|
|
43
|
-
if i < 80000:
|
|
44
|
-
|
|
45
|
-
s[0][i] = s1[i]
|
|
46
|
-
|
|
47
|
-
s[1][i] = s2[i]
|
|
48
|
-
|
|
49
|
-
s[0] = sp.zscore(s[0]) #原信号の正規化
|
|
50
|
-
|
|
51
|
-
s[1] = sp.zscore(s[1])
|
|
52
|
-
|
|
53
|
-
print("読み込み完了\nインパルス応答の作成\n...")
|
|
54
|
-
|
|
55
|
-
|
|
56
|
-
|
|
57
|
-
vc1 = np.convolve(s[0],A[0][0].real) + np.convolve(s[1],A[0][1].real)
|
|
58
|
-
|
|
59
|
-
vc2 = np.convolve(s[0],A[1][0].real) + np.convolve(s[1],A[1][1].real)
|
|
60
|
-
|
|
61
|
-
vc = np.array([[0 for i in range(2048)] for j in range(2)],dtype=complex)
|
|
62
|
-
|
|
63
|
-
for i in range(2048):
|
|
64
|
-
|
|
65
|
-
vc[0][i] = vc1[i]
|
|
66
|
-
|
|
67
|
-
vc[1][i] = vc2[i]
|
|
68
|
-
|
|
69
|
-
|
|
70
|
-
|
|
71
|
-
print("終了\nインパルス応答のフーリエ変換\n...")
|
|
72
|
-
|
|
73
|
-
for i in range(2):
|
|
74
|
-
|
|
75
|
-
for j in range(2):
|
|
76
|
-
|
|
77
|
-
A[i][j] = np.fft.fft(A[i][j])
|
|
78
|
-
|
|
79
|
-
|
|
80
|
-
|
|
81
|
-
print("終了\n短時間フーリエ変換\n...")
|
|
82
|
-
|
|
83
|
-
#frequency, segment time, STFT of x
|
|
84
|
-
|
|
85
|
-
frq1, t1, x1 = signal.stft(vc[0],fs=fs,nperseg=frame)
|
|
86
|
-
|
|
87
|
-
frq2, t2, x2 = signal.stft(vc[1],fs=fs,nperseg=frame)
|
|
88
|
-
|
|
89
|
-
|
|
90
|
-
|
|
91
|
-
fig = plt.figure()
|
|
92
|
-
|
|
93
|
-
ax1 = fig.add_subplot(4,1,1)
|
|
94
|
-
|
|
95
|
-
ax2 = fig.add_subplot(4,1,2)
|
|
96
|
-
|
|
97
|
-
#ax3 = fig.add_subplot(4,1,3)
|
|
98
|
-
|
|
99
|
-
#ax4 = fig.add_subplot(4,1,4)
|
|
100
|
-
|
|
101
|
-
ax1.plot(np.abs(x1))
|
|
102
|
-
|
|
103
|
-
ax2.plot(t1)
|
|
104
|
-
|
|
105
|
-
#ax3.plot(frq1)
|
|
106
|
-
|
|
107
|
-
#ax4.plot(x1[3])
|
|
108
|
-
|
|
109
|
-
fig.tight_layout()
|
|
110
|
-
|
|
111
|
-
plt.show()
|
|
112
|
-
|
|
113
|
-
```
|
|
114
|
-
|
|
115
25
|
|