''' '''
'''
ISC License
Copyright (c) 2016-2018, Autonomous Vehicle Systems Lab, University of Colorado at Boulder
Permission to use, copy, modify, and/or distribute this software for any
purpose with or without fee is hereby granted, provided that the above
copyright notice and this permission notice appear in all copies.
THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
'''
import math
import matplotlib.pyplot as plt
import numpy as np
from Basilisk.architecture import messaging
from Basilisk.fswAlgorithms import headingSuKF
from Basilisk.utilities import SimulationBaseClass, macros
import headingSuKF_test_utilities as FilterPlots
def setupFilterData(filterObject):
filterObject.alpha = 0.02
filterObject.beta = 2.0
filterObject.kappa = 0.0
# filterObject.state = [0.0, 0., 0., 0., 0.]
filterObject.stateInit = [0.0, 0.0, 1.0, 0.0, 0.0]
filterObject.covarInit = [0.1, 0.0, 0.0, 0.0, 0.0,
0.0, 0.1, 0.0, 0.0, 0.0,
0.0, 0.0, 0.1, 0.0, 0.0,
0.0, 0.0, 0.0, 0.01, 0.0,
0.0, 0.0, 0.0, 0.0, 0.01]
qNoiseIn = np.identity(5)
qNoiseIn[0:3, 0:3] = qNoiseIn[0:3, 0:3]*0.01*0.01
qNoiseIn[3:5, 3:5] = qNoiseIn[3:5, 3:5]*0.001*0.001
filterObject.qNoise = qNoiseIn.reshape(25).tolist()
[docs]
def test_functions_ukf(show_plots):
"""
**Validation Test Description**
This subtest runs through the general modules file for square root and unscented filters. These methods include
LU decompositions, QR decompositions that only provide the R matrix, as well as L, U inverses, and Cholesky
decompositions.
**Description of Variables Being Tested**
Each method of the general module file for square root and unscented filters are tested to machine precision
with errors of 1E-14
**General Documentation Comments**
This is a similar test used to other filter modules
"""
[testResults, testMessage] = heading_utilities_test(show_plots)
assert testResults < 1, testMessage
# uncomment this line is this test is to be skipped in the global unit test run, adjust message as needed
# @pytest.mark.skipif(conditionstring)
# uncomment this line if this test has an expected failure, adjust message as needed
# @pytest.mark.xfail() # need to update how the RW states are defined
# provide a unique test method name, starting with test_
[docs]
def test_all_heading_kf(show_plots):
"""
**Validation Test Description**
The StatePropSunLine subtest runs the filter and creates synthetic measurements to trigger the measurement update method.
This is tested in two parts. The filter first stabilizes to a value, and then the value is abruptly changed
in order for the filter to snap back to the solution. Measurements are provided every 10 seconds which provides the
sparse data that is usually characteristic of OpNav.
The StateUpdateSunLine subtest runs the filter without measurements to only trigger the time update method. This
ensures the filter stays at true values if no measurements are provided.
**Description of Variables Being Tested**
For the propagation: The state output by the filter is tested compared to the commanded target,
and the covariance is ensured to converge.
These are both tested to 1E-1 because of noise introduced in the measurements.
The measurement updated state output by the filter is tested compared to the expected target.
The stability of the state is tested to 1E-10.
"""
[testResults, testMessage] = StatePropSunLine(show_plots)
assert testResults < 1, testMessage
[testResults, testMessage] = StateUpdateSunLine(show_plots)
assert testResults < 1, testMessage
def heading_utilities_test(show_plots):
# The __tracebackhide__ setting influences pytest showing of tracebacks:
# the mrp_steering_tracking() function will not be shown unless the
# --fulltrace command line option is specified.
__tracebackhide__ = True
testFailCount = 0 # zero unit test result counter
testMessages = [] # create empty list to store test log messages
# Initialize the test module configuration data
AMatrix = [0.488894, 0.888396, 0.325191, 0.319207,
1.03469, -1.14707, -0.754928, 0.312859,
0.726885, -1.06887, 1.3703, -0.86488,
-0.303441, -0.809499, -1.71152, -0.0300513,
0.293871, -2.94428, -0.102242, -0.164879,
-0.787283, 1.43838, -0.241447, 0.627707]
RVector = headingSuKF.new_doubleArray(len(AMatrix))
AVector = headingSuKF.new_doubleArray(len(AMatrix))
for i in range(len(AMatrix)):
headingSuKF.doubleArray_setitem(AVector, i, AMatrix[i])
headingSuKF.doubleArray_setitem(RVector, i, 0.0)
headingSuKF.ukfQRDJustR(AVector, 6, 4, RVector)
RMatrix = []
for i in range(4*4):
RMatrix.append(headingSuKF.doubleArray_getitem(RVector, i))
RBaseNumpy = np.array(RMatrix).reshape(4,4)
AMatNumpy = np.array(AMatrix).reshape(6,4)
q,r = np.linalg.qr(AMatNumpy)
for i in range(r.shape[0]):
if r[i,i] < 0.0:
r[i,:] *= -1.0
if np.linalg.norm(r - RBaseNumpy) > 1.0E-15:
testFailCount += 1
testMessages.append("QR Decomposition accuracy failure")
AMatrix = [1.09327, 1.10927, -0.863653, 1.32288,
-1.21412, -1.1135, -0.00684933, -2.43508,
-0.769666, 0.371379, -0.225584, -1.76492,
-1.08906, 0.0325575, 0.552527, -1.6256,
1.54421, 0.0859311, -1.49159, 1.59683]
RVector = headingSuKF.new_doubleArray(len(AMatrix))
AVector = headingSuKF.new_doubleArray(len(AMatrix))
for i in range(len(AMatrix)):
headingSuKF.doubleArray_setitem(AVector, i, AMatrix[i])
headingSuKF.doubleArray_setitem(RVector, i, 0.0)
headingSuKF.ukfQRDJustR(AVector, 5, 4, RVector)
RMatrix = []
for i in range(4*4):
RMatrix.append(headingSuKF.doubleArray_getitem(RVector, i))
RBaseNumpy = np.array(RMatrix).reshape(4,4)
AMatNumpy = np.array(AMatrix).reshape(5,4)
q,r = np.linalg.qr(AMatNumpy)
for i in range(r.shape[0]):
if r[i,i] < 0.0:
r[i,:] *= -1.0
if np.linalg.norm(r - RBaseNumpy) > 1.0E-14:
testFailCount += 1
testMessages.append("QR Decomposition accuracy failure")
AMatrix = [ 0.2236, 0,
0, 0.2236,
-0.2236, 0,
0, -0.2236,
0.0170, 0,
0, 0.0170]
RVector = headingSuKF.new_doubleArray(len(AMatrix))
AVector = headingSuKF.new_doubleArray(len(AMatrix))
for i in range(len(AMatrix)):
headingSuKF.doubleArray_setitem(AVector, i, AMatrix[i])
headingSuKF.doubleArray_setitem(RVector, i, 0.0)
headingSuKF.ukfQRDJustR(AVector, 6, 2, RVector)
RMatrix = []
for i in range(2*2):
RMatrix.append(headingSuKF.doubleArray_getitem(RVector, i))
RBaseNumpy = np.array(RMatrix).reshape(2,2)
AMatNumpy = np.array(AMatrix).reshape(6,2)
q,r = np.linalg.qr(AMatNumpy)
for i in range(r.shape[0]):
if r[i,i] < 0.0:
r[i,:] *= -1.0
if np.linalg.norm(r - RBaseNumpy) > 1.0E-15:
testFailCount += 1
testMessages.append("QR Decomposition accuracy failure")
LUSourceMat = [8,1,6,3,5,7,4,9,2]
LUSVector = headingSuKF.new_doubleArray(len(LUSourceMat))
LVector = headingSuKF.new_doubleArray(len(LUSourceMat))
UVector = headingSuKF.new_doubleArray(len(LUSourceMat))
intSwapVector = headingSuKF.new_intArray(3)
for i in range(len(LUSourceMat)):
headingSuKF.doubleArray_setitem(LUSVector, i, LUSourceMat[i])
headingSuKF.doubleArray_setitem(UVector, i, 0.0)
headingSuKF.doubleArray_setitem(LVector, i, 0.0)
exCount = headingSuKF.ukfLUD(LUSVector, 3, 3, LVector, intSwapVector)
#headingSuKF.ukfUInv(LUSVector, 3, 3, UVector)
LMatrix = []
UMatrix = []
#UMatrix = []
for i in range(3):
currRow = headingSuKF.intArray_getitem(intSwapVector, i)
for j in range(3):
if(j<i):
LMatrix.append(headingSuKF.doubleArray_getitem(LVector, i*3+j))
UMatrix.append(0.0)
elif(j>i):
LMatrix.append(0.0)
UMatrix.append(headingSuKF.doubleArray_getitem(LVector, i*3+j))
else:
LMatrix.append(1.0)
UMatrix.append(headingSuKF.doubleArray_getitem(LVector, i*3+j))
# UMatrix.append(headingSuKF.doubleArray_getitem(UVector, i))
LMatrix = np.array(LMatrix).reshape(3,3)
UMatrix = np.array(UMatrix).reshape(3,3)
outMat = np.dot(LMatrix, UMatrix)
outMatSwap = np.zeros((3,3))
for i in range(3):
currRow = headingSuKF.intArray_getitem(intSwapVector, i)
outMatSwap[i,:] = outMat[currRow, :]
outMat[currRow,:] = outMat[i, :]
LuSourceArray = np.array(LUSourceMat).reshape(3,3)
if(np.linalg.norm(outMatSwap - LuSourceArray) > 1.0E-14):
testFailCount += 1
testMessages.append("LU Decomposition accuracy failure")
EqnSourceMat = [2.0, 1.0, 3.0, 2.0, 6.0, 8.0, 6.0, 8.0, 18.0]
BVector = [1.0, 3.0, 5.0]
EqnVector = headingSuKF.new_doubleArray(len(EqnSourceMat))
EqnBVector = headingSuKF.new_doubleArray(len(LUSourceMat)//3)
EqnOutVector = headingSuKF.new_doubleArray(len(LUSourceMat)//3)
for i in range(len(EqnSourceMat)):
headingSuKF.doubleArray_setitem(EqnVector, i, EqnSourceMat[i])
headingSuKF.doubleArray_setitem(EqnBVector, i//3, BVector[i//3])
headingSuKF.intArray_setitem(intSwapVector, i//3, 0)
headingSuKF.doubleArray_setitem(LVector, i, 0.0)
exCount = headingSuKF.ukfLUD(EqnVector, 3, 3, LVector, intSwapVector)
headingSuKF.ukfLUBckSlv(LVector, 3, 3, intSwapVector, EqnBVector, EqnOutVector)
expectedSol = [3.0/10.0, 4.0/10.0, 0.0]
errorVal = 0.0
for i in range(3):
errorVal += abs(headingSuKF.doubleArray_getitem(EqnOutVector, i) -expectedSol[i])
if(errorVal > 1.0E-14):
testFailCount += 1
testMessages.append("LU Back-Solve accuracy failure")
InvSourceMat = [8,1,6,3,5,7,4,9,2]
SourceVector = headingSuKF.new_doubleArray(len(InvSourceMat))
InvVector = headingSuKF.new_doubleArray(len(InvSourceMat))
for i in range(len(InvSourceMat)):
headingSuKF.doubleArray_setitem(SourceVector, i, InvSourceMat[i])
headingSuKF.doubleArray_setitem(InvVector, i, 0.0)
nRow = int(math.sqrt(len(InvSourceMat)))
headingSuKF.ukfMatInv(SourceVector, nRow, nRow, InvVector)
InvOut = []
for i in range(len(InvSourceMat)):
InvOut.append(headingSuKF.doubleArray_getitem(InvVector, i))
InvOut = np.array(InvOut).reshape(nRow, nRow)
expectIdent = np.dot(InvOut, np.array(InvSourceMat).reshape(3,3))
errorNorm = np.linalg.norm(expectIdent - np.identity(3))
if(errorNorm > 1.0E-14):
testFailCount += 1
testMessages.append("LU Matrix Inverse accuracy failure")
cholTestMat = [1.0, 0.0, 0.0, 0.0, 10.0, 5.0, 0.0, 5.0, 10.0]
SourceVector = headingSuKF.new_doubleArray(len(cholTestMat))
CholVector = headingSuKF.new_doubleArray(len(cholTestMat))
for i in range(len(cholTestMat)):
headingSuKF.doubleArray_setitem(SourceVector, i, cholTestMat[i])
headingSuKF.doubleArray_setitem(CholVector, i, 0.0)
nRow = int(math.sqrt(len(cholTestMat)))
headingSuKF.ukfCholDecomp(SourceVector, nRow, nRow, CholVector)
cholOut = []
for i in range(len(cholTestMat)):
cholOut.append(headingSuKF.doubleArray_getitem(CholVector, i))
cholOut = np.array(cholOut).reshape(nRow, nRow)
cholComp = np.linalg.cholesky(np.array(cholTestMat).reshape(nRow, nRow))
errorNorm = np.linalg.norm(cholOut - cholComp)
if(errorNorm > 1.0E-14):
testFailCount += 1
testMessages.append("Cholesky Matrix Decomposition accuracy failure")
InvSourceMat = [2.1950926119414667, 0.0, 0.0, 0.0,
1.0974804773131115, 1.9010439702743847, 0.0, 0.0,
0.0, 1.2672359635912551, 1.7923572711881284, 0.0,
1.0974804773131113, -0.63357997864171967, 1.7920348101787789, 0.033997451205364251]
SourceVector = headingSuKF.new_doubleArray(len(InvSourceMat))
InvVector = headingSuKF.new_doubleArray(len(InvSourceMat))
for i in range(len(InvSourceMat)):
headingSuKF.doubleArray_setitem(SourceVector, i, InvSourceMat[i])
headingSuKF.doubleArray_setitem(InvVector, i, 0.0)
nRow = int(math.sqrt(len(InvSourceMat)))
headingSuKF.ukfLInv(SourceVector, nRow, nRow, InvVector)
InvOut = []
for i in range(len(InvSourceMat)):
InvOut.append(headingSuKF.doubleArray_getitem(InvVector, i))
InvOut = np.array(InvOut).reshape(nRow, nRow)
expectIdent = np.dot(InvOut, np.array(InvSourceMat).reshape(nRow,nRow))
errorNorm = np.linalg.norm(expectIdent - np.identity(nRow))
if(errorNorm > 1.0E-12):
print(errorNorm)
testFailCount += 1
testMessages.append("L Matrix Inverse accuracy failure")
InvSourceMat = np.transpose(np.array(InvSourceMat).reshape(nRow, nRow)).reshape(nRow*nRow).tolist()
SourceVector = headingSuKF.new_doubleArray(len(InvSourceMat))
InvVector = headingSuKF.new_doubleArray(len(InvSourceMat))
for i in range(len(InvSourceMat)):
headingSuKF.doubleArray_setitem(SourceVector, i, InvSourceMat[i])
headingSuKF.doubleArray_setitem(InvVector, i, 0.0)
nRow = int(math.sqrt(len(InvSourceMat)))
headingSuKF.ukfUInv(SourceVector, nRow, nRow, InvVector)
InvOut = []
for i in range(len(InvSourceMat)):
InvOut.append(headingSuKF.doubleArray_getitem(InvVector, i))
InvOut = np.array(InvOut).reshape(nRow, nRow)
expectIdent = np.dot(InvOut, np.array(InvSourceMat).reshape(nRow,nRow))
errorNorm = np.linalg.norm(expectIdent - np.identity(nRow))
if(errorNorm > 1.0E-12):
print(errorNorm)
testFailCount += 1
testMessages.append("U Matrix Inverse accuracy failure")
# If the argument provided at commandline "--show_plots" evaluates as true,
# plot all figures
if show_plots:
plt.show()
# print out success message if no error were found
if testFailCount == 0:
print("PASSED: " + " UKF utilities")
# return fail count and join into a single string all messages in the list
# testMessage
return [testFailCount, ''.join(testMessages)]
def StateUpdateSunLine(show_plots):
__tracebackhide__ = True
testFailCount = 0 # zero unit test result counter
testMessages = [] # create empty list to store test log messages
unitTaskName = "unitTask" # arbitrary name (don't change)
unitProcessName = "TestProcess" # arbitrary name (don't change)
# Create a sim module as an empty container
unitTestSim = SimulationBaseClass.SimBaseClass()
# Create test thread
testProcessRate = macros.sec2nano(0.5) # update process rate update time
testProc = unitTestSim.CreateNewProcess(unitProcessName)
testProc.addTask(unitTestSim.CreateNewTask(unitTaskName, testProcessRate))
# Construct algorithm and associated C++ container
module = headingSuKF.headingSuKF()
module.ModelTag = "headingSuKF"
# Add test module to runtime call list
unitTestSim.AddModelToTask(unitTaskName, module)
setupFilterData(module)
dataLog = module.filtDataOutMsg.recorder()
unitTestSim.AddModelToTask(unitTaskName, dataLog)
testVector = np.array([0.9, 0.1, 0.02])
testOmega = np.array([0.01, 0.05, 0.001])
inputData = messaging.OpNavMsgPayload()
opnavDataInMsg = messaging.OpNavMsg()
stateTarget = testVector.tolist()
inputData.r_BN_B = stateTarget
stateTarget.extend([0.0, 0.0])
module.stateInit = [1., 0.2, 0.1, 0.01, 0.001]
# setup message connections
module.opnavDataInMsg.subscribeTo(opnavDataInMsg)
unitTestSim.InitializeSimulation()
t1 = 1000
for i in range(t1):
if i > 0 and i%20 == 0:
inputData.timeTag = macros.sec2nano(i * 0.5)
inputData.valid = 1
inputData.r_BN_B += np.random.normal(0, 0.001, 3)
inputData.covar_B = [0.0001**2, 0, 0, 0, 0.0001**2, 0, 0, 0, 0.0001**2]
opnavDataInMsg.write(inputData, unitTestSim.TotalSim.CurrentNanos)
unitTestSim.ConfigureStopTime(macros.sec2nano((i+1)*0.5))
unitTestSim.ExecuteSimulation()
stateLog = dataLog.state
postFitLog = dataLog.postFitRes
covarLog = dataLog.covar
stateTarget[:3] = (-testVector[:3]/np.linalg.norm(testVector[:3])).tolist()
for i in range(5):
# check covariance immediately after measurement is taken,
# ensure order of magnitude less than initial covariance.
if(np.abs(covarLog[t1-10, i*5+i] - covarLog[0, i*5+i]/10) > 1E-1):
testFailCount += 1
testMessages.append("Covariance update failure")
if(abs(stateLog[-1, i] - stateTarget[i-1]) > 1.0E-1):
print(abs(stateLog[-1, i] - stateTarget[i-1]))
testFailCount += 1
testMessages.append("State update failure")
testVector = np.array([0.6, -0.1, 0.2])
stateTarget = testVector.tolist()
inputData.r_BN_B = stateTarget
stateTarget.extend([0.0, 0.0])
for i in range(t1):
if i%20 == 0:
inputData.timeTag = macros.sec2nano(i*0.5 +t1 +1)
inputData.r_BN_B += np.random.normal(0, 0.001, 3)
inputData.valid = 1
inputData.covar_B = [0.0001**2,0,0,0,0.0001**2,0,0,0,0.0001**2]
opnavDataInMsg.write(inputData, unitTestSim.TotalSim.CurrentNanos)
unitTestSim.ConfigureStopTime(macros.sec2nano((i+t1 +1)*0.5))
unitTestSim.ExecuteSimulation()
stateLog = dataLog.state
stateErrorLog = dataLog.stateError
postFitLog = dataLog.postFitRes
covarLog = dataLog.covar
stateTarget[:3] = (-testVector[:3]/np.linalg.norm(testVector[:3])).tolist()
for i in range(5):
if(np.abs(covarLog[2*t1-10, i*5+i] - covarLog[0, i*5+i]/10)>1E-1):
testFailCount += 1
testMessages.append("Covariance update failure")
if(abs(stateLog[-1, i] - stateTarget[i-1]) > 1.0E-1):
print(abs(stateLog[-1, i] - stateTarget[i-1]))
testFailCount += 1
testMessages.append("State update failure")
FilterPlots.StateCovarPlot(dataLog.times(), stateLog, covarLog, 'Update', show_plots)
FilterPlots.StateCovarPlot(dataLog.times(),stateErrorLog, covarLog, 'Update_Error', show_plots)
FilterPlots.PostFitResiduals(dataLog.times(), postFitLog, 0.001, 'Update', show_plots)
# print out success message if no error were found
if testFailCount == 0:
print("PASSED: " + module.ModelTag + " state update")
# return fail count and join into a single string all messages in the list
# testMessage
return [testFailCount, ''.join(testMessages)]
def StatePropSunLine(show_plots):
__tracebackhide__ = True
testFailCount = 0 # zero unit test result counter
testMessages = [] # create empty list to store test log messages
unitTaskName = "unitTask" # arbitrary name (don't change)
unitProcessName = "TestProcess" # arbitrary name (don't change)
# Create a sim module as an empty container
unitTestSim = SimulationBaseClass.SimBaseClass()
# Create test thread
testProcessRate = macros.sec2nano(0.5) # update process rate update time
testProc = unitTestSim.CreateNewProcess(unitProcessName)
testProc.addTask(unitTestSim.CreateNewTask(unitTaskName, testProcessRate))
# Construct algorithm and associated C++ container
module = headingSuKF.headingSuKF()
module.ModelTag = "headingSuKF"
# Add test module to runtime call list
unitTestSim.AddModelToTask(unitTaskName, module)
setupFilterData(module)
dataLog = module.filtDataOutMsg.recorder()
unitTestSim.AddModelToTask(unitTaskName, dataLog)
inData = messaging.OpNavMsgPayload()
inDataMsg = messaging.OpNavMsg().write(inData)
# setup message connections
module.opnavDataInMsg.subscribeTo(inDataMsg)
unitTestSim.InitializeSimulation()
unitTestSim.ConfigureStopTime(macros.sec2nano(8000.0))
unitTestSim.ExecuteSimulation()
stateLog = dataLog.state
postFitLog = dataLog.postFitRes
covarLog = dataLog.covar
FilterPlots.StateCovarPlot(dataLog.times(), stateLog, covarLog, 'Prop', show_plots)
FilterPlots.PostFitResiduals(dataLog.times(), postFitLog, module.qObsVal, 'Prop', show_plots)
for i in range(5):
if(abs(stateLog[-1, i] - stateLog[0, i]) > 1.0E-10):
print(abs(stateLog[-1, i] - stateLog[0, i]))
testFailCount += 1
testMessages.append("State propagation failure")
# print out success message if no error were found
if testFailCount == 0:
print("PASSED: " + module.ModelTag + " state propagation")
# return fail count and join into a single string all messages in the list
# testMessage
return [testFailCount, ''.join(testMessages)]
if __name__ == "__main__":
# test_all_heading_kf(True)
StateUpdateSunLine(True)
# StatePropSunLine(True)