def hilbert(data: str) -> dict:
signal = VibrationSignal(
data=binary_deserializer(data), fs=10000, compute_axis=False
)
env = signal.to_envelope()
env.compute_spectrum(compute_axis=False)
return {"vib": signal.data, "env_vib": env.data, "env_fft": env.spec}
async def read_warning_log_data_by_id(data_id: int,
mp_id: int,
fault_mode: FaultRule,
conn: Database = Depends(get_db)):
"""
Get warning log data by ID and fault pattern.
"""
data = await get_data_by_id(
conn=conn,
mp_id=mp_id,
orm_model=VibData,
require_mp_type=0,
data_id=data_id,
)
if not data:
raise HTTPException(status_code=400, detail="Item not found")
signal = VibrationSignal(data=np.fromstring(data["ima"], dtype=np.float32),
fs=10000,
type=2)
if (fault_mode == FaultRule.ub or fault_mode == FaultRule.ma
or fault_mode == FaultRule.al or fault_mode == FaultRule.bl
or fault_mode == FaultRule.rb):
signal = signal.to_velocity(detrend_type="poly")
signal.compute_spectrum()
return {"spec": signal.spec, "freq": signal.freq}
if fault_mode == FaultRule.bw:
signal = signal.to_filted_signal(filter_type="highpass",
co_frequency=2 * 1000 /
10000).to_envelope()
signal.compute_spectrum()
return {"spec": signal.spec, "freq": signal.freq}
if fault_mode == FaultRule.sg:
signal.compute_spectrum()
return {"spec": signal.spec, "freq": signal.freq}
def acceleration_to_velocity(data: str) -> dict:
acc = VibrationSignal(data=binary_deserializer(data), fs=10000)
vel = acc.to_velocity()
acc.compute_spectrum(compute_axis=False)
vel.compute_spectrum(compute_axis=False)
return {
"vel": vel.data,
"acc": acc.data,
"acc_spec": acc.spec,
"vel_spec": vel.spec,
}
def surge_diagnosis(self, diag_obj: VibrationSignal):
low_index1, low_energy1 = diag_obj.get_band_energy(
fr=self.fr, band_range=(0, 0.4)
)
low_index2, low_energy2 = diag_obj.get_band_energy(
fr=self.fr, band_range=(0.6, 0.8)
)
self.sg_index = np.array([low_index1, low_index2])
self.lt_fr_amp = np.array([low_energy1, low_energy2])
self.sg_indicator = 0.45 * low_energy1 + 0.35 * low_energy2
self.sg_level = np.searchsorted(self.sg_threshold, self.sg_indicator)
def calculate_feature_row(row, station_id, inner_station_id):
feature_model = VibFeature.model(station_id=station_id, inner_id=inner_station_id)
signal = np.fromstring(row.ima, dtype=np.float32)
signal = VibrationSignal(data=signal, fs=10000)
return feature_model(
rms=float(signal.rms_fea),
max=float(signal.max_fea),
p2p=float(signal.pp_fea),
avg=float(np.mean(signal.data)),
var=float(signal.var_fea),
kurtosis=float(signal.kurtosis),
data_id=row.id,
time=row.time,
)
def fast_fournier_transform(signal: str) -> dict:
signal = VibrationSignal(
data=binary_deserializer(signal), fs=10000, compute_axis=False
)
signal.compute_spectrum(compute_axis=False)
return {"spec": signal.spec, "vib": signal.data}
def empirical_mode_decomposition(data: str) -> dict:
signal = VibrationSignal(
data=binary_deserializer(data), fs=10000, compute_axis=False
)
res = signal.get_empirical_mode_decomposition()
return res
def welch_spectrum_estimation(data: str) -> dict:
signal = VibrationSignal(
data=binary_deserializer(data), fs=10000, compute_axis=False
)
res = signal.get_welch_spectrum_estimation()
return {"vib": signal.data, **res}
def multi_scale_envelope_spectrum(data: str) -> dict:
signal = VibrationSignal(
data=binary_deserializer(data), fs=10000, compute_axis=False
)
res = signal.get_multi_scale_envelope_spectrum(n_Ssta=1.0, n_Send=8.0, n_Sint=0.2)
return res
def short_time_fournier_transform(data: str) -> dict:
signal = VibrationSignal(
data=binary_deserializer(data), fs=10000, compute_axis=False
)
res = signal.get_short_time_fournier_transform()
return res