def test_serialization4():
from siconos.kernel import LagrangianLinearTIDS, NewtonImpactNSL, \
LagrangianLinearTIR, Interaction, Model, MoreauJeanOSI, TimeDiscretisation, LCP, TimeStepping
from numpy import array, eye, empty
t0 = 0 # start time
T = 10 # end time
h = 0.005 # time step
r = 0.1 # ball radius
g = 9.81 # gravity
m = 1 # ball mass
e = 0.9 # restitution coeficient
theta = 0.5 # theta scheme
#
# dynamical system
#
x = array([1, 0, 0]) # initial position
v = array([0, 0, 0]) # initial velocity
mass = eye(3) # mass matrix
mass[2, 2] = 3./5 * r * r
# the dynamical system
ball = LagrangianLinearTIDS(x, v, mass)
# set external forces
weight = array([-m * g, 0, 0])
ball.setFExtPtr(weight)
#
# Interactions
#
# ball-floor
H = array([[1, 0, 0]])
nslaw = NewtonImpactNSL(e)
relation = LagrangianLinearTIR(H)
inter = Interaction(1, nslaw, relation)
#
# Model
#
first_bouncingBall = Model(t0, T)
# add the dynamical system to the non smooth dynamical system
first_bouncingBall.nonSmoothDynamicalSystem().insertDynamicalSystem(ball)
# link the interaction and the dynamical system
first_bouncingBall.nonSmoothDynamicalSystem().link(inter, ball)
#
# Simulation
#
# (1) OneStepIntegrators
OSI = MoreauJeanOSI(theta)
# (2) Time discretisation --
t = TimeDiscretisation(t0, h)
# (3) one step non smooth problem
osnspb = LCP()
# (4) Simulation setup with (1) (2) (3)
s = TimeStepping(t)
s.insertIntegrator(OSI)
s.insertNonSmoothProblem(osnspb)
# end of model definition
#
# computation
#
# simulation initialization
first_bouncingBall.setSimulation(s)
first_bouncingBall.initialize()
#
# save and load data from xml and .dat
#
from siconos.io.io_base import save, load
save(first_bouncingBall, "bouncingBall.xml")
bouncingBall = load("bouncingBall.xml")
# the number of time steps
N = (T-t0)/h+1
# Get the values to be plotted
# ->saved in a matrix dataPlot
dataPlot = empty((N, 5))
#
# numpy pointers on dense Siconos vectors
#
q = ball.q()
v = ball.velocity()
p = ball.p(1)
lambda_ = inter.lambda_(1)
#
# initial data
#
dataPlot[0, 0] = t0
dataPlot[0, 1] = q[0]
dataPlot[0, 2] = v[0]
dataPlot[0, 3] = p[0]
dataPlot[0, 4] = lambda_[0]
k = 1
# time loop
while(s.hasNextEvent()):
s.computeOneStep()
dataPlot[k, 0] = s.nextTime()
dataPlot[k, 1] = q[0]
dataPlot[k, 2] = v[0]
dataPlot[k, 3] = p[0]
dataPlot[k, 4] = lambda_[0]
k += 1
print(s.nextTime())
s.nextStep()
#
# comparison with the reference file
#
from siconos.kernel import SimpleMatrix, getMatrix
from numpy.linalg import norm
ref = getMatrix(SimpleMatrix(os.path.join(working_dir,
"data/result.ref")))
assert (norm(dataPlot - ref) < 1e-12)
myProcessRelation = ZhuravlevTwistingR(C, B) myNslaw = RelayNSL(2) myNslaw.display() myProcessInteraction = Interaction(myNslaw, myProcessRelation) filippov = NonSmoothDynamicalSystem(t0, T) filippov.insertDynamicalSystem(process) filippov.link(myProcessInteraction, process) td = TimeDiscretisation(t0, h) s = TimeStepping(filippov, td) myIntegrator = EulerMoreauOSI(theta) s.insertIntegrator(myIntegrator) #TODO python <- SICONOS_RELAY_LEMKE # access dparam osnspb = Relay() s.insertNonSmoothProblem(osnspb) s.setComputeResiduY(True) s.setComputeResiduR(True) s.setNewtonMaxIteration(30) s.setNewtonTolerance(1e-14) # matrix to save data dataPlot = empty((N + 1, 5)) control = empty((N + 1, )) dataPlot[0, 0] = t0
broadphase = SiconosBulletCollisionManager() # insert a non smooth law for contactors id 0 broadphase.insertNonSmoothLaw(nslaw, 0, 0) # The ground is a static object # we give it a group contactor id : 0 scs = SiconosContactorSet() scs.append(SiconosContactor(ground)) broadphase.insertStaticContactorSet(scs, groundOffset) # (6) Simulation setup with (1) (2) (3) (4) (5) simulation = TimeStepping(timedisc) simulation.insertInteractionManager(broadphase) simulation.insertIntegrator(osi) simulation.insertNonSmoothProblem(osnspb) # simulation initialization bouncingBox.setSimulation(simulation) bouncingBox.initialize() # Get the values to be plotted # ->saved in a matrix dataPlot N = int((T - t0) / h) dataPlot = zeros((N + 1, 4)) # # numpy pointers on dense Siconos vectors
# Declaration of the Dynamical System
processDS = FirstOrderLinearDS(x0, A)
processDS.setComputebFunction("RelayPlugin", "computeB")
# Model
process = Model(t0, T)
process.nonSmoothDynamicalSystem().insertDynamicalSystem(processDS)
# time discretisation
processTD = TimeDiscretisation(t0, h)
tSensor = TimeDiscretisation(t0, hControl)
tActuator = TimeDiscretisation(t0, hControl)
# Creation of the Simulation
processSimulation = TimeStepping(processTD, 0)
processSimulation.setName("plant simulation")
# Declaration of the integrator
processIntegrator = ZeroOrderHoldOSI(processDS)
processSimulation.insertIntegrator(processIntegrator)
# Actuator, Sensor & ControlManager
control = ControlManager(process)
sens = LinearSensor(tSensor, processDS, sensorC)
control.addSensorPtr(sens)
act = LinearSMCOT2(tActuator, processDS)
act.setCsurfacePtr(Csurface)
act.addSensorPtr(sens)
control.addActuatorPtr(act)
# Initialization
process.initialize(processSimulation)
control.initialize()
# This is not working right now
#eventsManager = s.eventsManager()
def test_bouncing_ball1():
from siconos.kernel import LagrangianLinearTIDS, NewtonImpactNSL, \
LagrangianLinearTIR, Interaction, Model, MoreauJeanOSI, TimeDiscretisation, LCP, TimeStepping
from numpy import array, eye, empty
t0 = 0 # start time
T = 10 # end time
h = 0.005 # time step
r = 0.1 # ball radius
g = 9.81 # gravity
m = 1 # ball mass
e = 0.9 # restitution coeficient
theta = 0.5 # theta scheme
#
# dynamical system
#
x = array([1, 0, 0]) # initial position
v = array([0, 0, 0]) # initial velocity
mass = eye(3) # mass matrix
mass[2, 2] = 3./5 * r * r
# the dynamical system
ball = LagrangianLinearTIDS(x, v, mass)
# set external forces
weight = array([-m * g, 0, 0])
ball.setFExtPtr(weight)
#
# Interactions
#
# ball-floor
H = array([[1, 0, 0]])
nslaw = NewtonImpactNSL(e)
relation = LagrangianLinearTIR(H)
inter = Interaction(1, nslaw, relation)
#
# Model
#
bouncingBall = Model(t0, T)
# add the dynamical system to the non smooth dynamical system
bouncingBall.nonSmoothDynamicalSystem().insertDynamicalSystem(ball)
# link the interaction and the dynamical system
bouncingBall.nonSmoothDynamicalSystem().link(inter, ball)
#
# Simulation
#
# (1) OneStepIntegrators
OSI = MoreauJeanOSI(theta)
OSI.insertDynamicalSystem(ball)
# (2) Time discretisation --
t = TimeDiscretisation(t0, h)
# (3) one step non smooth problem
osnspb = LCP()
# (4) Simulation setup with (1) (2) (3)
s = TimeStepping(t)
s.insertIntegrator(OSI)
s.insertNonSmoothProblem(osnspb)
# end of model definition
#
# computation
#
# simulation initialization
bouncingBall.initialize(s)
#
# save and load data from xml and .dat
#
try:
from siconos.io import save
save(bouncingBall, "bouncingBall.xml")
save(bouncingBall, "bouncingBall.bin")
except:
print("Warning : could not import save from siconos.io")
# the number of time steps
N = (T-t0)/h+1
# Get the values to be plotted
# ->saved in a matrix dataPlot
dataPlot = empty((N, 5))
#
# numpy pointers on dense Siconos vectors
#
q = ball.q()
v = ball.velocity()
p = ball.p(1)
lambda_ = inter.lambda_(1)
#
# initial data
#
dataPlot[0, 0] = t0
dataPlot[0, 1] = q[0]
dataPlot[0, 2] = v[0]
dataPlot[0, 3] = p[0]
dataPlot[0, 4] = lambda_[0]
k = 1
# time loop
while(s.hasNextEvent()):
s.computeOneStep()
dataPlot[k, 0] = s.nextTime()
dataPlot[k, 1] = q[0]
dataPlot[k, 2] = v[0]
dataPlot[k, 3] = p[0]
dataPlot[k, 4] = lambda_[0]
k += 1
#print(s.nextTime())
s.nextStep()
#
# comparison with the reference file
#
from siconos.kernel import SimpleMatrix, getMatrix
from numpy.linalg import norm
ref = getMatrix(SimpleMatrix(os.path.join(working_dir, "data/result.ref")))
assert (norm(dataPlot - ref) < 1e-12)
myProcessInteraction = Interaction(ninter, myNslaw,
myProcessRelation)
myNSDS = NonSmoothDynamicalSystem()
myNSDS.insertDynamicalSystem(process)
myNSDS.link(myProcessInteraction,process)
filippov = Model(t0,T)
filippov.setNonSmoothDynamicalSystemPtr(myNSDS)
td = TimeDiscretisation(t0, h)
s = TimeStepping(td)
myIntegrator = EulerMoreauOSI(process, theta)
s.insertIntegrator(myIntegrator)
#TODO python <- SICONOS_RELAY_LEMKE
# access dparam
osnspb = Relay()
s.insertNonSmoothProblem(osnspb)
s.setComputeResiduY(True)
s.setComputeResiduR(True)
filippov.initialize(s);
# matrix to save data
dataPlot = empty((N+1,5))
control = empty((N+1,))
# Simulation # # (1) OneStepIntegrators OSI = MoreauJeanOSI(theta) OSI.insertDynamicalSystem(ball) # (2) Time discretisation -- t = TimeDiscretisation(t0, h) # (3) one step non smooth problem osnspb = LCP() # (4) Simulation setup with (1) (2) (3) s = TimeStepping(t) s.insertIntegrator(OSI) s.insertNonSmoothProblem(osnspb) # end of model definition # # computation # # simulation initialization bouncingBall.initialize(s) # the number of time steps N = (T - t0) / h # Get the values to be plotted
# (1) OneStepIntegrators OSI = MoreauJeanOSI(theta) # (2) Time discretisation -- t = TimeDiscretisation(t0, h) # (3) one step non smooth problem osnspb = LCP() # (4) Simulation setup with (1) (2) (3) s = TimeStepping(t) #s.setDisplayNewtonConvergence(True) s.setNewtonTolerance(1e-10) #s.setNewtonMaxIteration(1) s.insertIntegrator(OSI) s.insertNonSmoothProblem(osnspb) model.setSimulation(s) # end of model definition # # computation # # simulation initialization model.initialize() # Get the values to be plotted # ->saved in a matrix dataPlot dataPlot = np.empty((N + 1, 26)) #
def test_smc1():
from siconos.kernel import FirstOrderLinearDS, Model, TimeDiscretisation, \
TimeStepping, ZeroOrderHoldOSI, TD_EVENT
from siconos.control.simulation import ControlManager
from siconos.control.sensor import LinearSensor
from siconos.control.controller import LinearSMCOT2
from numpy import eye, empty, zeros
import numpy as np
from math import ceil, sin
# Derive our own version of FirstOrderLinearDS
class MyFOLDS(FirstOrderLinearDS):
def computeb(self, time):
t = sin(50*time)
# XXX fix this !
u = [t, -t]
self.setb(u)
# variable declaration
ndof = 2 # Number of degrees of freedom of your system
t0 = 0.0 # start time
T = 1 # end time
h = 1.0e-4 # time step for simulation
hControl = 1.0e-2 # time step for control
Xinit = 1.0 # initial position
N = ceil((T-t0)/h + 10) # number of time steps
outputSize = 4 # number of variable to store at each time step
# Matrix declaration
A = zeros((ndof, ndof))
x0 = [Xinit, -Xinit]
Brel = np.array([[0], [1]])
sensorC = eye(ndof)
sensorD = zeros((ndof, ndof))
Csurface = [[0, 1.0]]
# Simple check
if h > hControl:
print("hControl must be bigger than h")
exit(1)
# Declaration of the Dynamical System
processDS = MyFOLDS(x0, A)
# XXX b is not automatically created ...
# processDS.setb([0, 0])
# Model
process = Model(t0, T)
process.nonSmoothDynamicalSystem().insertDynamicalSystem(processDS)
# time discretization
processTD = TimeDiscretisation(t0, h)
tSensor = TimeDiscretisation(t0, hControl)
tActuator = TimeDiscretisation(t0, hControl)
# Creation of the Simulation
processSimulation = TimeStepping(processTD, 0)
processSimulation.setName("plant simulation")
# Declaration of the integrator
processIntegrator = ZeroOrderHoldOSI()
process.nonSmoothDynamicalSystem().setOSI(processDS, processIntegrator)
processSimulation.insertIntegrator(processIntegrator)
# Actuator, Sensor & ControlManager
control = ControlManager(processSimulation)
sens = LinearSensor(processDS, sensorC, sensorD)
control.addSensorPtr(sens, tSensor)
act = LinearSMCOT2(sens)
act.setCsurface(Csurface)
act.setB(Brel)
control.addActuatorPtr(act, tActuator)
# Initialization.
process.initialize(processSimulation)
control.initialize(process)
# This is not working right now
# eventsManager = s.eventsManager()
# Matrix for data storage
dataPlot = empty((3*(N+1), outputSize))
dataPlot[0, 0] = t0
dataPlot[0, 1] = processDS.x()[0]
dataPlot[0, 2] = processDS.x()[1]
dataPlot[0, 3] = act.u()[0]
# Main loop
k = 1
while processSimulation.hasNextEvent():
if processSimulation.eventsManager().nextEvent().getType() == TD_EVENT:
processSimulation.computeOneStep()
dataPlot[k, 0] = processSimulation.nextTime()
dataPlot[k, 1] = processDS.x()[0]
dataPlot[k, 2] = processDS.x()[1]
dataPlot[k, 3] = act.u()[0]
k += 1
processSimulation.nextStep()
# print processSimulation.nextTime()
# Resize matrix
dataPlot.resize(k, outputSize)