def compute_raim_position(gps_week, gps_sow, prns, prns_pos, pranges,
bcestore):
if len(prns) == 0 or len(prns_pos) == 0:
return np.array([0, 0, 0])
t = gpstk.GPSWeekSecond(gps_week, gps_sow).toCommonTime()
prnList = [gpstk.SatID(int(i[3:])) for i in prns]
satVector = gpstk.seqToVector(list(prnList), outtype='vector_SatID')
rangeVector = gpstk.seqToVector([float(i) for i in pranges])
noTropModel = gpstk.ZeroTropModel()
raimSolver = gpstk.PRSolution2()
raimSolver.RAIMCompute(t, satVector, rangeVector, bcestore, noTropModel)
r = np.array([
raimSolver.Solution[0], raimSolver.Solution[1], raimSolver.Solution[2]
])
return r
def main():
parser = argparse.ArgumentParser()
parser.add_argument('rinex3obs_filename')
parser.add_argument('rinex3nav_filename')
parser.add_argument('-m', '--rinexmet_filename')
args = parser.parse_args()
# Declaration of objects for storing ephemerides and handling RAIM
bcestore = gpstk.GPSEphemerisStore()
raimSolver = gpstk.PRSolution2()
# Object for void-type tropospheric model (in case no meteorological
# RINEX is available)
noTropModel = gpstk.ZeroTropModel()
# Object for GG-type tropospheric model (Goad and Goodman, 1974)
# Default constructor => default values for model
ggTropModel = gpstk.GGTropModel()
# Pointer to one of the two available tropospheric models. It points
# to the void model by default
tropModel = noTropModel
navHeader, navData = gpstk.readRinex3Nav(args.rinex3nav_filename)
for navDataObj in navData:
ephem = navDataObj.toEngEphemeris()
bcestore.addEphemeris(ephem)
# Setting the criteria for looking up ephemeris:
bcestore.SearchNear()
if args.rinexmet_filename is not None:
metHeader, metData = gpstk.readRinexMet(args.rinexmet_filename)
tropModel = ggTropModel
obsHeader, obsData = gpstk.readRinex3Obs(args.rinex3obs_filename)
# The following lines fetch the corresponding indexes for some
# observation types we are interested in. Given that old-style
# observation types are used, GPS is assumed.
try:
indexP1 = obsHeader.getObsIndex('P1')
except:
print 'The observation files has no P1 pseudoranges.'
sys.exit()
try:
indexP2 = obsHeader.getObsIndex('P2')
except:
indexP2 = -1
for obsObj in obsData:
# Find a weather point. Only if a meteorological RINEX file
# was provided, the meteorological data linked list "rml" is
# neither empty or at its end, and the time of meteorological
# records are below observation data epoch.
if args.rinexmet_filename is not None:
for metObj in metData:
if metObj.time >= obsObj.time:
break
else:
metDataDict = metObj.getData()
temp = metDataDict[gpstk.RinexMetHeader.TD]
pressure = metDataDict[gpstk.RinexMetHeader.PR]
humidity = metDataDict[gpstk.RinexMetHeader.HR]
ggTropModel.setWeather(temp, pressure, humidity)
if obsObj.epochFlag == 0 or obsObj.epochFlag == 1:
# Note that we use lists here, but we will need types backed
# by C++ std::vectors later. We'll just keep it easy and use
# gpstk.seqToVector to convert them. If there was a speed
# bottleneck we could use gpstk.cpp.vector_SatID and
# gpstk.cpp.vector_double though.
prnList = []
rangeList = []
# This part gets the PRN numbers and ionosphere-corrected
# pseudoranges for the current epoch. They are correspondly fed
# into "prnList" and "rangeList"; "obs" is a public attribute of
# Rinex3ObsData to get the map of observations
for satID, datumList in obsObj.obs.iteritems():
# The RINEX file may have P1 observations, but the current
# satellite may not have them.
P1 = 0.0
try:
P1 = obsObj.getObs(satID, indexP1).data
except gpstk.exceptions.Exception:
continue # Ignore this satellite if P1 is not found
ionocorr = 0.0
# If there are P2 observations, let's try to apply the
# ionospheric corrections
if indexP2 >= 0:
# The RINEX file may have P2 observations, but the
# current satellite may not have them.
P2 = 0.0
try:
P2 = obsObj.getObs(satID, indexP2).data
except gpstk.exceptions.Exception:
continue # Ignore this satellite if P1 is not found
# list 'vecList' contains RinexDatum, whose public
# attribute "data" indeed holds the actual data point
ionocorr = 1.0 / (1.0 - gpstk.constants.GAMMA_GPS) * (P1 -
P2)
# Now, we include the current PRN number in the first part
# of "it" iterator into the list holding the satellites.
# All satellites in view at this epoch that have P1 or P1+P2
# observations will be included.
prnList.append(satID)
# The same is done for the list of doubles holding the
# corrected ranges
rangeList.append(P1 - ionocorr)
# WARNING: Please note that so far no further correction
# is done on data: Relativistic effects, tropospheric
# correction, instrumental delays, etc
# The default constructor for PRSolution2 objects (like
# "raimSolver") is to set a RMSLimit of 6.5. We change that
# here. With this value of 3e6 the solution will have a lot
# more dispersion.
raimSolver.RMSLimit = 3e6
# In order to compute positions we need the current time, the
# vector of visible satellites, the vector of corresponding
# ranges, the object containing satellite ephemerides, and a
# pointer to the tropospheric model to be applied
time = obsObj.time
# the RAIMComputer method of PRSolution2 accepts a vector<SatID> as its
# 2nd argument, but the list is of RinexSatID, which is a subclass of SatID.
# Since C++ containers are NOT covariant, it is neccessary to change the
# output to a vector or SatID's rather thta a vector of RinexSatID's.
satVector = gpstk.cpp.seqToVector(prnList, outtype='vector_SatID')
rangeVector = gpstk.cpp.seqToVector(rangeList)
raimSolver.RAIMCompute(time, satVector, rangeVector, bcestore,
tropModel)
# Note: Given that the default constructor sets public
# attribute "Algebraic" to FALSE, a linearized least squares
# algorithm will be used to get the solutions.
# Also, the default constructor sets ResidualCriterion to true,
# so the rejection criterion is based on RMS residual of fit,
# instead of RMS distance from an a priori position.
if raimSolver.isValid():
# Vector "Solution" holds the coordinates, expressed in
# meters in an Earth Centered, Earth Fixed (ECEF) reference
# frame. The order is x, y, z (as all ECEF objects)
x, y, z = raimSolver.Solution[0], raimSolver.Solution[
1], raimSolver.Solution[2]
print "%12.5f %12.5f %12.5f" % (x, y, z)
navfn = gpstk.getPathData() + '/arlm200z.15n'
metfn = gpstk.getPathData() + '/arlm200z.15m'
navHeader, navData = gpstk.readRinex3Nav(navfn)
ephStore = gpstk.gpstk.Rinex3EphemerisStore()
for navDataObj in navData:
ephStore.addEphemeris(navDataObj)
tropModel = gpstk.GGTropModel()
metHeader, metData = gpstk.readRinexMet(metfn)
obsHeader, obsData = gpstk.readRinex3Obs(obsfn)
indexP1 = obsHeader.getObsIndex('C1W')
indexP2 = obsHeader.getObsIndex('C2W')
raimSolver = gpstk.PRSolution2()
for obsObj in obsData:
for metObj in metData:
if metObj.time >= obsObj.time:
break
else:
metDataDict = metObj.getData()
temp = metDataDict[gpstk.RinexMetHeader.TD]
pressure = metDataDict[gpstk.RinexMetHeader.PR]
humidity = metDataDict[gpstk.RinexMetHeader.HR]
tropModel.setWeather(temp, pressure, humidity)
if obsObj.epochFlag == 0 or obsObj.epochFlag == 1:
# Note that we use lists here, but we will need types backed
# by C++ std::vectors later. We'll just keep it easy and use