-
Notifications
You must be signed in to change notification settings - Fork 0
/
process_flight_data.py
299 lines (255 loc) · 15.3 KB
/
process_flight_data.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
import math
import numpy as np
import pandas as pd
from bokeh.io import export_png
from bokeh.models import HoverTool
from bokeh.models.formatters import DatetimeTickFormatter
from bokeh.plotting import figure, show
from scipy.signal import savgol_filter
# local imports
from interpolate import interpolate
# lunar results
# Force/torque sensor data
L20_d4_a = pd.read_csv('Flight data/20201002 131424 sliptest s20-p1500-d4.csv')
L20_d4_b = pd.read_csv('Flight data/20201002 132114 sliptest s20-p1500-d4.csv')
L20_d4_c = pd.read_csv('Flight data/20201002 135545 sliptest s20-p1500-d4.csv')
L20_d4_d = pd.read_csv('Flight data/20201002 140110 sliptest s20-p1500-d4.csv')
L70_d4_a = pd.read_csv('Flight data/20201002 134029 sliptest s70-p1500-d4.csv')
L70_d4_b = pd.read_csv('Flight data/20201002 134618 sliptest s70-p1500-d4.csv')
L70_d4_c = pd.read_csv('Flight data/20201002 135133 sliptest s70-p1500-d4.csv')
# trio data (contains motor current and sinkage)
L20_d4_a_s = pd.read_csv('Flight data/20201002-131424-trio-s20-p1500-d4_combined.CSV')
L20_d4_b_s = pd.read_csv('Flight data/20201002-132114-trio-s20-p1500-d4_combined.CSV')
L20_d4_c_s = pd.read_csv('Flight data/20201002-135545-trio-s20-p1500-d4_combined.CSV')
L20_d4_d_s = pd.read_csv('Flight data/20201002-140110-trio-s20-p1500-d4_combined.CSV')
L70_d4_a_s = pd.read_csv('Flight data/20201002-134029-trio-s70-p1500-d4_combined.CSV')
L70_d4_b_s = pd.read_csv('Flight data/20201002-134618-trio-s70-p1500-d4_combined.CSV')
L70_d4_c_s = pd.read_csv('Flight data/20201002-135133-trio-s70-p1500-d4_combined.CSV')
# make some lists
slip20_files = [L20_d4_a, L20_d4_b, L20_d4_c, L20_d4_d]
slip70_files = [L70_d4_a, L70_d4_b, L70_d4_c]
all_files = [slip20_files, slip70_files]
slip20_trio_files = [L20_d4_a_s, L20_d4_b_s, L20_d4_c_s, L20_d4_d_s]
slip70_trio_files = [L70_d4_a_s, L70_d4_b_s, L70_d4_c_s]
all_trio_files = [slip20_trio_files, slip70_trio_files]
# Force/torque sensor offsets
x_offset = 13
y_offset = 9.8
z_offset = -12
# Indices where motion starts
starttimes_slip20_index = [23850, 19802, 19098, 19741]
starttimes_slip70_index = [34498, 27681, 24328]
endtimes_slip20_index = np.array(starttimes_slip20_index) + 7000
endtimes_slip70_index = np.array(starttimes_slip70_index) + 7000
starttimes_slip20 = [25172, 21160, 20406, 21068]
starttimes_slip70 = [35850, 29074, 25694]
all_starttimes = [starttimes_slip20_index, starttimes_slip70_index]
endtimes_slip20 = [29167, 25151, 24334, 25049]
endtimes_slip70 = [39788, 33232, 29836]
all_endtimes = [endtimes_slip20, endtimes_slip70]
# measured soil slopes (for slope correction)
slopes_slip20 = np.array([[6.48, 8.79, 6.81, 8.28], # measured soil angle at beginning,
[4.68, 5.91, 7.97, 6.39], # middle, and
[4.1, 4.2, 5.78, 3.88]]) # end of each test
slopes_slip70 = np.array([[6.04, 7.2, 6.24], # measured soil angle at beginning,
[7.54, 8.1, 6.05], # middle, and
[6.26, 6.82, 4.87]]) # end of each test
# measured soil heights (for slope correction)
delta_y_slip20 = np.array([21.47, 38.65, 44.48, 33.82])
delta_y_slip70 = np.array([13.04, 13.04, 10.14])
slip20_soilheight = np.array([[7.43, 4.25, 2.48, 4.25, 5.66],
[6.73, 9.56, 7.79, 7.08, 3.54],
[8.5, 9.2, 7.79, 9.91, 8.5],
[6.02, 9.56, 7.08, 5.31, 14.16]])
slip70_soilheight = np.array([[3.54, 4.25, 4.6, 2.83, 8.85],
[4.25, 4.25, 3.19, 1.77, 3.54],
[3.89, 1.42, 2.48, 1.06, 1.06]])
all_slopes = [slopes_slip20, slopes_slip70]
# process data
for i in range(0, len(all_files)):
files = all_files[i]
trio_files = all_trio_files[i]
starttimes = all_starttimes[i]
endtimes = all_endtimes[i]
slopes = all_slopes[i]
for ii in range(0, len(files)):
# set dataframe indices
files[ii].set_index(pd.DatetimeIndex(files[ii]['timestamp']), inplace=True)
trio_files[ii].set_index(pd.DatetimeIndex(trio_files[ii]['timestamp']), inplace=True)
# convert indices to elapsed seconds
files[ii].index = files[ii].index - files[ii].index[0]
trio_files[ii].index = trio_files[ii].index - trio_files[ii].index[0]
# correct for FTS offsets
files[ii].force_x = -files[ii].force_x + x_offset
files[ii].force_y = -files[ii].force_y + y_offset
files[ii].force_z = -files[ii].force_z + z_offset
# friction correction
files[ii].force_x = files[ii].force_x + 0.25 * files[ii].force_z
# slope correction for the drawbar pull data
length = len(files[ii].force_x[starttimes[ii]:endtimes[ii]])
midpoint = math.floor(length / 2)
first_half = np.linspace(slopes[0, ii], slopes[1, ii], midpoint)
if length % 2 == 0:
second_half = np.linspace(slopes[1, ii], slopes[2, ii], midpoint)
else:
second_half = np.linspace(slopes[1, ii], slopes[2, ii], midpoint + 1)
slope_angles = np.zeros(length)
slope_angles[0:midpoint] = first_half
slope_angles[midpoint:length] = second_half
files[ii].force_x[starttimes[ii]:endtimes[ii]] = \
files[ii].force_x[starttimes[ii]:endtimes[ii]] - \
files[ii].force_y[starttimes[ii]:endtimes[ii]] * \
np.cos(np.radians(slope_angles)) * np.sin(np.radians(slope_angles))
# smooth
files[ii].force_x = savgol_filter(files[ii].force_x, 41, 2) # window 41, order 2
files[ii].force_y = savgol_filter(files[ii].force_y, 41, 2) # window 41, order 2
# convert potentiometer counts to mm of sinkage and smooth
trio_files[ii].sinkage = savgol_filter(trio_files[ii].sinkage, 201, 2) # window size 41, polynomial order 2
trio_files[ii].sinkage = (trio_files[ii].sinkage - 110) / 132.24
trio_files[ii].sinkage = trio_files[ii].sinkage - trio_files[ii].sinkage[0]
# smooth the current data
trio_files[ii].maxon_current = savgol_filter(trio_files[ii].maxon_current, 41, 2) # window 41, order 2
# plot DP/W
# 20% slip in blue, 70% slip in red.
ylabel = 'DP/W'
plot3 = figure(toolbar_location=None, x_axis_label='Elapsed Time (s)', y_axis_label=ylabel, x_range=(1, 27000), y_range=(0, 0.65))
#
for j in range(0, len(slip20_files)):
plot3.line(slip20_files[j].index[starttimes_slip20[j]:endtimes_slip20[j]] -
slip20_files[j].index[starttimes_slip20_index[j]],
slip20_files[j].force_x[starttimes_slip20[j]:endtimes_slip20[j]] /
slip20_files[j].force_y[starttimes_slip20[j]:endtimes_slip20[j]],
legend="20% slip")
#
for k in range(0, len(slip70_files)):
plot3.line(slip70_files[k].index[starttimes_slip70[k]:endtimes_slip70[k]] -
slip70_files[k].index[starttimes_slip70_index[k]],
slip70_files[k].force_x[starttimes_slip70[k]:endtimes_slip70[k]] /
slip70_files[k].force_y[starttimes_slip70[k]:endtimes_slip70[k]],
line_color="red", line_dash='dashed', legend="70% slip")
# format plot and export to png
plot3.xaxis.formatter = DatetimeTickFormatter(minsec=['%M:%S'])
hover_tool = HoverTool()
plot3.tools.append(hover_tool)
plot3.add_tools(HoverTool(show_arrow=False, line_policy='next', tooltips=[
("Time (ms)", "$x"),
("Value", "$y")
], formatters=dict(x='datetime')))
plot3.xaxis.axis_label_text_font_size = "14pt"
plot3.xaxis.major_label_text_font_size = "14pt"
plot3.yaxis.axis_label_text_font_size = "14pt"
plot3.yaxis.major_label_text_font_size = "14pt"
plot3.legend.label_text_font_size = "14pt"
show(plot3)
export_png(plot3, filename="plots/drawbar-pull-lunar.png")
# plot sinkage
ylabel = 'Sinkage (mm)'
plot = figure(toolbar_location=None, x_axis_label='Elapsed Time (s)', y_axis_label=ylabel, x_range=(1, 27000), y_range=(-2, 31))
for j in range(0, len(slip20_trio_files)):
# apply slope correction to sinkage data
# get angle at each point (same procedure used for FTS data, applied to trio time stamps)
length = len(slip20_trio_files[j].sinkage)
# make vector with soil height values from 0-5s, 5-10s, 10-15s, 15-20s, and 20s-end:
delta_y = np.zeros(length)
delta_y[0:500] = np.linspace(0, slip20_soilheight[j, 0], 500)
delta_y[500:1000] = np.linspace(0, slip20_soilheight[j, 1], 500) + slip20_soilheight[j, 0]
delta_y[1000:1500] = np.linspace(0, slip20_soilheight[j, 2], 500) + slip20_soilheight[j, 1]\
+ slip20_soilheight[j, 0]
delta_y[1500:2000] = np.linspace(0, slip20_soilheight[j, 3], 500) + slip20_soilheight[j, 2]\
+ slip20_soilheight[j, 1] + slip20_soilheight[j, 0]
delta_y[2000:length] = np.linspace(0, slip20_soilheight[j, 4], length-2000) + slip20_soilheight[j, 3] \
+ slip20_soilheight[j, 2] + slip20_soilheight[j, 1] + slip20_soilheight[j, 0]
# plot sinkage data with slope correction
plot.line(slip20_trio_files[j].index[0:length-500], slip20_trio_files[j].sinkage[0:length-500] -
delta_y[0:length-500], legend="20% slip")
# set up dataframes to export to .csv
fn = pd.DataFrame(slip20_files[j].force_y[starttimes_slip20_index[j]:endtimes_slip20[j]])
fn.set_index(slip20_files[j].index[starttimes_slip20_index[j]:endtimes_slip20[j]] -
slip20_files[j].index[starttimes_slip20_index[j]], inplace=True)
dp = pd.DataFrame(slip20_files[j].force_x[starttimes_slip20_index[j]:endtimes_slip20[j]])
dp.set_index(slip20_files[j].index[starttimes_slip20_index[j]:endtimes_slip20[j]] -
slip20_files[j].index[starttimes_slip20_index[j]], inplace=True)
dp_raw = pd.DataFrame(slip20_files[j].force_x[starttimes_slip20_index[j]:endtimes_slip20[j]] - 0.25 * \
slip20_files[j].force_x[starttimes_slip20_index[j]:endtimes_slip20[j]])
dp_raw.set_index(slip20_files[j].index[starttimes_slip20_index[j]:endtimes_slip20[j]] -
slip20_files[j].index[starttimes_slip20_index[j]], inplace=True)
torque = pd.DataFrame(slip20_files[j].torque_z[starttimes_slip20_index[j]:endtimes_slip20[j]])
torque.set_index(slip20_files[j].index[starttimes_slip20_index[j]:endtimes_slip20[j]] -
slip20_files[j].index[starttimes_slip20_index[j]], inplace=True)
# interpolate sinkage and current data to the timestamps of the force/torque sensor data
sink = interpolate(slip20_trio_files[j].index[0:length - 500] - slip20_trio_files[j].index[0], fn.index,
slip20_trio_files[j].sinkage[0:length - 500] - delta_y[0:length - 500])
sink = pd.DataFrame(sink)
sink.set_index(fn.index, inplace=True)
current = interpolate(slip20_trio_files[j].index[0:length - 500] - slip20_trio_files[j].index[0], fn.index,
slip20_trio_files[j].maxon_current[0:length - 500])
current = pd.DataFrame(current / 1000)
current.set_index(fn.index, inplace=True)
# export processed timeseries data to csv
size = np.amin(np.array([fn.size, dp.size, sink.size]))
timeseries_data = pd.concat([dp[0:size], fn[0:size], sink[0:size], torque[0:size], dp_raw[0:size], current[0:size]]
, axis=1)
timeseries_data.columns = ['dp', 'fn', 'z', 'torque', 'dp_raw', 'current']
timeseries_data.to_csv('processed data/lunar-g_20-%s.csv' % j, index=True, index_label='Timestamp')
for k in range(0, len(slip70_trio_files)):
# apply slope correction to sinkage data
# get angle at each point (same procedure used for FTS data, applied to trio time stamps)
length = len(slip70_trio_files[k].sinkage)
# make vector with soil height values from 0-5s, 5-10s, 10-15s, 15-20s, and 20s-end:
delta_y = np.zeros(length)
delta_y[0:500] = np.linspace(0, slip70_soilheight[k, 0], 500)
delta_y[500:1000] = np.linspace(0, slip70_soilheight[k, 1], 500) + slip70_soilheight[k, 0]
delta_y[1000:1500] = np.linspace(0, slip70_soilheight[k, 2], 500) + slip70_soilheight[k, 1] \
+ slip70_soilheight[k, 0]
delta_y[1500:2000] = np.linspace(0, slip70_soilheight[k, 3], 500) + slip70_soilheight[k, 2] \
+ slip70_soilheight[k, 1] + slip70_soilheight[k, 0]
delta_y[2000:length] = np.linspace(0, slip70_soilheight[k, 3], length - 2000) + slip70_soilheight[k, 3] \
+ slip70_soilheight[k, 2] + slip70_soilheight[k, 1] + slip70_soilheight[k, 0]
# plot sinkage data with slope correction
plot.line(slip70_trio_files[k].index[0:length-500], slip70_trio_files[k].sinkage[0:length-500] -
delta_y[0:length-500], line_color="red",
line_dash='dashed', legend="70% slip")
# set up dataframes to export to .csv
fn = pd.DataFrame(slip70_files[k].force_y[starttimes_slip70_index[k]:endtimes_slip70[k]])
fn.set_index(slip70_files[k].index[starttimes_slip70_index[k]:endtimes_slip70[k]] -
slip70_files[k].index[starttimes_slip70_index[k]], inplace=True)
dp = pd.DataFrame(slip70_files[k].force_x[starttimes_slip70_index[k]:endtimes_slip70[k]])
dp.set_index(slip70_files[k].index[starttimes_slip70_index[k]:endtimes_slip70[k]] -
slip70_files[k].index[starttimes_slip70_index[k]], inplace=True)
dp_raw = pd.DataFrame(slip70_files[k].force_x[starttimes_slip70_index[k]:endtimes_slip70[k]] - 0.25 * \
slip70_files[k].force_x[starttimes_slip70_index[k]:endtimes_slip70[k]])
dp_raw.set_index(slip70_files[k].index[starttimes_slip70_index[k]:endtimes_slip70[k]] -
slip70_files[k].index[starttimes_slip70_index[k]], inplace=True)
torque = pd.DataFrame(slip70_files[k].torque_z[starttimes_slip70_index[k]:endtimes_slip70[k]])
torque.set_index(slip70_files[k].index[starttimes_slip70_index[k]:endtimes_slip70[k]] -
slip70_files[k].index[starttimes_slip70_index[k]], inplace=True)
# interpolate sinkage and current data to the timestamps of the force/torque sensor data
sink = interpolate(slip70_trio_files[k].index[0:length - 500] - slip70_trio_files[k].index[0], fn.index,
slip70_trio_files[k].sinkage[0:length - 500] - delta_y[0:length - 500])
sink = pd.DataFrame(sink)
sink.set_index(fn.index, inplace=True)
current = interpolate(slip70_trio_files[k].index[0:length - 500] - slip70_trio_files[k].index[0], fn.index,
slip70_trio_files[k].maxon_current[0:length - 500])
current = pd.DataFrame(current / 1000)
current.set_index(fn.index, inplace=True)
# export processed timeseries data to csv
size = np.amin(np.array([fn.size, dp.size, sink.size]))
timeseries_data = pd.concat([dp[0:size], fn[0:size], sink[0:size], torque[0:size], dp_raw[0:size], current[0:size]],
axis=1)
timeseries_data.columns = ['dp', 'fn', 'z', 'torque', 'dp_raw', 'current']
timeseries_data.to_csv('processed data/lunar-g_70-%s.csv' % k, index=True, index_label='Timestamp')
# format plot and export to png
plot.xaxis.formatter = DatetimeTickFormatter(minsec=['%M:%S'])
hover_tool = HoverTool()
plot.tools.append(hover_tool)
plot.add_tools(HoverTool(show_arrow=False, line_policy='next', tooltips=[
("Time (ms)", "$x"),
("Value", "$y")
], formatters=dict(x='datetime')))
plot.xaxis.axis_label_text_font_size = "14pt"
plot.xaxis.major_label_text_font_size = "14pt"
plot.yaxis.axis_label_text_font_size = "14pt"
plot.yaxis.major_label_text_font_size = "14pt"
plot.legend.label_text_font_size = "14pt"
show(plot)
export_png(plot, filename="plots/sinkage-lunar.png")