Ground Truth about 10_nlos_correction_ HOT 22 CLOSED

WLS positioning result.
NLOS exclusion available 9.0 9.0 0.0
meanEr_ 46.16285547 std_---- 29.6756165967
percent_1= 0.21387283237 percent_2= 0.450867052023 percent_3= 0.410404624277
meanErPros_ 44.3619354996 stdProp_ 29.9814039684
percent_1= 0.277456647399 percent_2= 0.49710982659 percent_3= 0.398843930636

def PercentCal(self, data_):
# self.axes_1.grid(True)
error_ = []
error_ = data_
percent_1 = 0.0 # >15
percent_2 = 0.0 # >25
percent_3 = 0.0 # >40
for perCal in range(len(error_)):
if (error_[perCal] <= 20):
percent_1 = percent_1 + 1
if (error_[perCal] <= 35):
percent_2 = percent_2 + 1
if (error_[perCal] >= 45):
percent_3 = percent_3 + 1
percent_1 = percent_1 / len(error_)
percent_2 = percent_2 / len(error_)
percent_3 = percent_3 / len(error_)
print 'percent_1=', percent_1, 'percent_2=', percent_2, 'percent_3=', percent_3

from 10_nlos_correction_.

from 10_nlos_correction_.

self.excluSatLis = [9,14,31]

from 10_nlos_correction_.

Surface Detection

from 10_nlos_correction_.

NLOS exclusion available 9.0 8.0 1.0
meanEr_ 46.16285547 std_---- 29.6756165967
percent_1= 0.21387283237 percent_2= 0.450867052023 percent_3= 0.410404624277
meanErPros_ 43.9014069661 stdProp_ 29.4982060128
percent_1= 0.248554913295 percent_2= 0.531791907514 percent_3= 0.393063583815

from 10_nlos_correction_.

Covariance estimation with NLOS modelling

from 10_nlos_correction_.

Covariance estimation with NLOS modelling: add adjustment coefficient

from 10_nlos_correction_.

(correction1 + userError) * HDOP

from 10_nlos_correction_.

(correction1 +correction1 + userError) * HDOP

from 10_nlos_correction_.

(correction1 +correction1 + userError) * HDOP ( Coefficient : 0.35)

from 10_nlos_correction_.

(correction1 +correction1 + userError) * HDOP ( Coefficient : 0.43)

from 10_nlos_correction_.

(correction1 +correction1 + userError) * HDOP ( Coefficient : 0.43 distance=8)

from 10_nlos_correction_.

from 10_nlos_correction_.

(correction1 +correction1 + userError) * HDOP ( Coefficient : 0.43 distance=5.8)

from 10_nlos_correction_.

self.axesCurv_3.plot(self.GNSSTimeProp, self.errorProp, '-', linewidth='1', color='black',
label='errorProp')
self.axesCurv_3.plot(self.GNSSTimeProp, self.Covariance, '-', linewidth='1', color='blue',
label='Covariance') # self.Covariance
# CovarianceFixed
self.axesCurv_3.plot(self.GNSSTimeProp, self.CovarianceFixed, '-*', linewidth='1', color='Red',
label='Covariance') # self.Covariance

from 10_nlos_correction_.

from 10_nlos_correction_.

LiDAR localization: covariance <40

from 10_nlos_correction_.

from 10_nlos_correction_.

from 10_nlos_correction_.

constant

Fixed Covariance (13) and add all the GNSS positioning into the graph optimization

from 10_nlos_correction_.

Conventional method:
--- pose graph optimization ---
nodes: 207 edges: 515
optimizing... done
iterations: 17
chi2: (before)2.11194e+15 -> (after)11.8739
time: 0.000[sec]

from 10_nlos_correction_.

--- pose graph optimization ---
nodes: 195 edges: 384
optimizing... done
iterations: 45
chi2: (before)-nan -> (after)-nan
time: 0.211[sec]
enu_():6.19839
-38.175

from 10_nlos_correction_.