Moving Horizon Estimation
Main.MovingHorizonEstimation History
Hide minor edits - Show changes to output
Changed lines 7-8 from:
The basic idea of MHE is to minimize the discrepancy between the measured outputs of the system and the outputs predicted by a model of the system, subject to a set of constraints on the states of the system. This optimization problem can be written as follows:
to:
The basic idea of MHE is to minimize the discrepancy between the measured outputs of the system and the outputs predicted by a model of the system, subject to a set of constraints on the states of the system. This optimization problem is often written as follows:
Changed line 17 from:
where {xk,...,xk+N} are the states of the system at each time step in the horizon, ''y'_i_''' is the measured output at time ''i'', ''h(x'_i_')'' is the model prediction of the output at time ''i'', ''f(x'_i_')'' is the model prediction of the state at time ''i+1'' given the state at time ''i'', ''x'_k,0_''' is the initial estimate of the state at time ''k'', and ''g'_j_'(x'_i_')'' are the constraints on the states at time ''i''.
to:
where {xk,...,xk+N} are the states of the system at each time step in the horizon, ''y'_i_''' is the measured output at time ''i'', ''h(x'_i_')'' is the model prediction of the output at time ''i'', ''f(x'_i_')'' is the model prediction of the state at time ''i+1'' given the state at time ''i'', ''x'_k,0_''' is the initial estimate of the state at time ''k'', and ''g'_j_'(x'_i_')'' are the constraints on the states at time ''i''. Additional optimization forms are discussed in [[Main/EstimatorObjective|Estimator Objectives]].
Changed lines 15-17 from:
{$g(x_i) \leq 0 \quad (j=1,2,...,m)$}
where {xk,...,xk+N} are the states of the system at each time step in the horizon, ''y'_i_''' is the measured output at time ''i'', ''h(x'_i_')'' is the model prediction of the output at time ''i'', ''f(x'_i_')'' is the model prediction of the state at time ''i+1'' given the state at time ''i'', ''x'_k,0_''' is the initial estimate of the state at time ''k'', and ''g(x'_i_')'' are the constraints on the states at time ''i''.
where {xk,...,xk+N} are the states of the system at each time step in the horizon, ''y'_i_''' is the measured output at time ''i'', ''h(x'_i_')'' is the model prediction of the output at time ''i'', ''f(x'_i_')'' is the model prediction of the state at time ''i+1'' given the state at time ''i'', ''x'_k,0_''' is the initial estimate of the state at time ''k'', and ''g(x'_i_')'' are the constraints on the states at time ''i''.
to:
{$g_j(x_i) \leq 0 \quad (j=1,2,...,m)$}
where {xk,...,xk+N} are the states of the system at each time step in the horizon, ''y'_i_''' is the measured output at time ''i'', ''h(x'_i_')'' is the model prediction of the output at time ''i'', ''f(x'_i_')'' is the model prediction of the state at time ''i+1'' given the state at time ''i'', ''x'_k,0_''' is the initial estimate of the state at time ''k'', and ''g'_j_'(x'_i_')'' are the constraints on the states at time ''i''.
where {xk,...,xk+N} are the states of the system at each time step in the horizon, ''y'_i_''' is the measured output at time ''i'', ''h(x'_i_')'' is the model prediction of the output at time ''i'', ''f(x'_i_')'' is the model prediction of the state at time ''i+1'' given the state at time ''i'', ''x'_k,0_''' is the initial estimate of the state at time ''k'', and ''g'_j_'(x'_i_')'' are the constraints on the states at time ''i''.
Changed lines 15-17 from:
{$g_j(x_i) \leq 0 \quad (j=1,2,...,m)$}
where {xk,...,xk+N} are the states of the system at each time step in the horizon, ''y'_i_''' is the measured output at time ''i'', ''h(x'_i_')'' is the model prediction of the output at time ''i'', ''f(x'_i_')'' is the model prediction of the state at time ''i+1'' given the state at time ''i'', ''x'_k,0_''' is the initial estimate of the state at time ''k'', and ''g'_j_'(x'_i_')'' are the constraints on the states at time ''i''.
where {xk,...,xk+N} are the states of the system at each time step in the horizon, ''y'_i_''' is the measured output at time ''i'', ''h(x'_i_')'' is the model prediction of the output at time ''i'', ''f(x'_i_')'' is the model prediction of the state at time ''i+1'' given the state at time ''i'', ''x'_k,0_''' is the initial estimate of the state at time ''k'', and ''g
to:
{$g(x_i) \leq 0 \quad (j=1,2,...,m)$}
where {xk,...,xk+N} are the states of the system at each time step in the horizon, ''y'_i_''' is the measured output at time ''i'', ''h(x'_i_')'' is the model prediction of the output at time ''i'', ''f(x'_i_')'' is the model prediction of the state at time ''i+1'' given the state at time ''i'', ''x'_k,0_''' is the initial estimate of the state at time ''k'', and ''g(x'_i_')'' are the constraints on the states at time ''i''.
where {xk,...,xk+N} are the states of the system at each time step in the horizon, ''y'_i_''' is the measured output at time ''i'', ''h(x'_i_')'' is the model prediction of the output at time ''i'', ''f(x'_i_')'' is the model prediction of the state at time ''i+1'' given the state at time ''i'', ''x'_k,0_''' is the initial estimate of the state at time ''k'', and ''g(x'_i_')'' are the constraints on the states at time ''i''.
Changed line 17 from:
where {xk,...,xk+N} are the states of the system at each time step in the horizon, ''y'_i_''' is the measured output at time ''i'', ''h(x'_i_')'' is the model prediction of the output at time ''i'', ''f(x'_i_')'' is the model prediction of the state at time ''i+1'' given the state at time ''i'', {x'_k,0_''' is the initial estimate of the state at time ''k'', and ''g_j(x'_i_')'' are the constraints on the states at time ''i''.
to:
where {xk,...,xk+N} are the states of the system at each time step in the horizon, ''y'_i_''' is the measured output at time ''i'', ''h(x'_i_')'' is the model prediction of the output at time ''i'', ''f(x'_i_')'' is the model prediction of the state at time ''i+1'' given the state at time ''i'', ''x'_k,0_''' is the initial estimate of the state at time ''k'', and ''g'_j_'(x'_i_')'' are the constraints on the states at time ''i''.
Added lines 4-23:
Moving horizon estimation (MHE) is a type of optimization-based model predictive control algorithm that is used to estimate the states of a system based on noisy measurements. It is an online algorithm that operates in a rolling horizon fashion, meaning that it continually updates the estimates of the states as new measurements become available.
The basic idea of MHE is to minimize the discrepancy between the measured outputs of the system and the outputs predicted by a model of the system, subject to a set of constraints on the states of the system. This optimization problem can be written as follows:
{$\min_{x_k, ..., x_{k+N}} \sum_{i=k}^{k+N} ||y_i - h(x_i)||^2 + \sum_{i=k}^{k+N-1} ||x_{i+1} - f(x_i)||^2$}
subject to:
{$x_k = x_{k,0}$}
{$g_j(x_i) \leq 0 \quad (j=1,2,...,m)$}
where {xk,...,xk+N} are the states of the system at each time step in the horizon, ''y'_i_''' is the measured output at time ''i'', ''h(x'_i_')'' is the model prediction of the output at time ''i'', ''f(x'_i_')'' is the model prediction of the state at time ''i+1'' given the state at time ''i'', {x'_k,0_''' is the initial estimate of the state at time ''k'', and ''g_j(x'_i_')'' are the constraints on the states at time ''i''.
The optimization problem is solved using an optimization algorithm, such as the interior point method, at each time step as new measurements become available. The solution of the optimization problem gives the estimates of the states of the system for the current time step and the next few time steps in the horizon. The horizon is then shifted forward by one time step and the process is repeated.
MHE can be useful for systems that are subject to noise or disturbance, as it allows the model to adapt to these changes and maintain accurate estimates of the states. It is also useful for systems with constraints on the states, as the constraints can be incorporated into the optimization problem.
----
Changed line 205 from:
plt.plot(time,Tc_meas,'k-',linewidth=2)
to:
plt.plot(time,Tc_meas,'k-',lw=2)
Changed line 212 from:
plt.plot(time,UA_mhe_store,'r:',linewidth=2)
to:
plt.plot(time,UA_mhe_store,'r:',lw=2)
Changed line 219 from:
plt.plot(time,T_mhe_store,'b-',linewidth=2)
to:
plt.plot(time,T_mhe_store,'b-',lw=2)
Changed line 226 from:
plt.plot(time,Ca_mhe_store,'m-',linewidth=2)
to:
plt.plot(time,Ca_mhe_store,'m-',lw=2)
Deleted lines 195-200:
if i==0:
# remove '.value' overrides to use '.MEAS'
for xi in [Ca,T,Tc]:
xi.value=[None]*len(s.time)
Tc_mhe.value=[None]*len(m.time)
Added line 36:
# Moving Horizon Estimation
Changed lines 49-53 from:
#%% Simulation
s = GEKKO(name=
#1 step of simulation, discretization matches MHE
s = GEKKO(
#
to:
# Simulation
s = GEKKO(remote=False,name='cstr-sim')
# One step of simulation, discretization matches MHE
s = GEKKO(remote=False,name='cstr-sim')
# One step of simulation, discretization matches MHE
Changed line 55 from:
#Receive measurement from simulated control
to:
# Receive measurement from simulated control
Changed line 60 from:
#State variables to watch
to:
# Simulator variables
Changed lines 62-64 from:
T = s.SV(value=305,lb=250,ub=500,name='t')
#other parameters
#
to:
T = s.SV(value=335,lb=250,ub=500,name='t')
# Parameters
# Parameters
Changed lines 71-72 from:
k0 = s.Param(value=7.2*10**10)
UA = s.Param(value=5*10**4)
UA = s.Param(value=
to:
k0 = s.Param(value=7.2e10)
UA = s.Param(value=5e4)
UA = s.Param(value=5e4)
Changed line 76 from:
#Variables
to:
# Variables
Changed line 80 from:
#Rate equations
to:
# Rate equations
Changed line 83 from:
#CSTR equations
to:
# CSTR equations
Changed line 87 from:
#Options
to:
# Options
Changed lines 93-107 from:
#%% MHE
#Model
m = GEKKO(name=
#6 time points in horizon
m.time = np.linspace(0,.5,6)
#Parameter to Estimate
UA_mhe = mFV(value=10*10**4,name='ua')
UA_mhe.STATUS = 1 #estimate
UA_mhe.FSTATUS = 0 #no measurements
#upper and lower bounds for optimizer
UA_mhe.LOWER = 10000
#Model
m = GEKKO(
#
#Parameter to Estimate
UA_mhe = m
UA_mhe
#upper and lower bounds for optimizer
UA_mhe.
to:
# MHE
m = GEKKO(remote=False,name='cstr-mhe')
# 11 time points in horizon (0.1 each step)
m.time = np.linspace(0,1.0,11)
# Parameter to Estimate
UA_mhe = m.FV(value=1e4,name='ua')
UA_mhe.STATUS = 1 # estimate
UA_mhe.FSTATUS = 0 # no measurements
# Upper and lower bounds for optimizer
UA_mhe.LOWER = 30000
m = GEKKO(remote=False,name='cstr-mhe')
# 11 time points in horizon (0.1 each step)
m.time = np.linspace(0,1.0,11)
# Parameter to Estimate
UA_mhe = m.FV(value=1e4,name='ua')
UA_mhe.STATUS = 1 # estimate
UA_mhe.FSTATUS = 0 # no measurements
# Upper and lower bounds for optimizer
UA_mhe.LOWER = 30000
Changed line 107 from:
#Measurement input
to:
# Cooling Jacket Temperature
Changed lines 109-121 from:
Tc_mhe.STATUS = 0 #don't estimate
Tc_mhe.FSTATUS = 1 #receive measurement
#Measurement to match simulation with
T_mhe =m.CV(value=325 ,lb=250,ub=500,name='t')
T_mhe.STATUS = 1 #minimize error between simulation and measurement
T_mhe.FSTATUS = 1 #receive measurement
T_mhe.MEAS_GAP = 0.1 #measurement deadband gap
#State to watch
Ca_mhe =m.SV(value=0.8, ub=1, lb=0,name='ca')
#Other parameters
Tc_mhe.FSTATUS = 1 #receive measurement
#
T_mhe =
T_mhe.
T_mhe.
#State to watch
Ca_mhe =
#
to:
Tc_mhe.STATUS = 0 # don't estimate
Tc_mhe.FSTATUS = 1 # receive measurement
# Reactor Temperature
T_mhe = m.CV(value=335,lb=250,ub=500,name='t')
T_mhe.FSTATUS = 1 # minimize error with measurement
T_mhe.MEAS_GAP = 0.1 # measurement deadband gap
# Reactor Concentration
Ca_mhe = m.SV(value=0.5, ub=1, lb=0,name='ca')
# Parameters
Tc_mhe.FSTATUS = 1 # receive measurement
# Reactor Temperature
T_mhe = m.CV(value=335,lb=250,ub=500,name='t')
T_mhe.FSTATUS = 1 # minimize error with measurement
T_mhe.MEAS_GAP = 0.1 # measurement deadband gap
# Reactor Concentration
Ca_mhe = m.SV(value=0.5, ub=1, lb=0,name='ca')
# Parameters
Changed line 127 from:
k0 = m.Param(value=7.2*10**10)
to:
k0 = m.Param(value=7.2e10)
Changed line 131 from:
#Equation variables(2 other DOF from CV and FV)
to:
# Equation variables (2 other DOF from CV and FV)
Changed line 135 from:
#Reaction equations
to:
# Reaction equations
Changed lines 138-144 from:
#CSTR equations
m.Equation(V* Ca_mhe.dt() == q*(Ca0-Ca_mhe)-V*rate) #mol balance
m.Equation(rho*Cp*V* T_mhe.dt() == q*rho*Cp*(T0-T_mhe)+ V*mdelH*rate + UA_mhe*(Tc_mhe-T_mhe)) #energy balance
#Global Tuning
m.options.IMODE = 5 #MHE
m.Equation(V* Ca_mhe.dt() == q*(Ca0-Ca_mhe)-V*rate) #mol balance
m.Equation(rho*Cp*V* T_mhe.dt() == q*rho*Cp*(T0-T_mhe)
#Global Tuning
m.options.IMODE = 5
to:
# CSTR equations
m.Equation(V* Ca_mhe.dt() == q*(Ca0-Ca_mhe)-V*rate) # mol balance
m.Equation(rho*Cp*V* T_mhe.dt() == q*rho*Cp*(T0-T_mhe) \
+ V*mdelH*rate + UA_mhe*(Tc_mhe-T_mhe)) # energy balance
# Global Tuning
m.options.IMODE = 5 # MHE
m.Equation(V* Ca_mhe.dt() == q*(Ca0-Ca_mhe)-V*rate) # mol balance
m.Equation(rho*Cp*V* T_mhe.dt() == q*rho*Cp*(T0-T_mhe) \
+ V*mdelH*rate + UA_mhe*(Tc_mhe-T_mhe)) # energy balance
# Global Tuning
m.options.IMODE = 5 # MHE
Changed lines 147-151 from:
m.options.SOLVER = 3 #IPOPT
#%% Loop
# number of cycles to run
#
# number of cycles
to:
m.options.SOLVER = 3 # IPOPT
# Cycles to run
# Cycles to run
Changed lines 157-158 from:
time = np.linspace(0,cycles*dt-dt,cycles) # time points for plot
to:
# time points for plot
time = np.linspace(0,cycles*dt-dt,cycles)
time = np.linspace(0,cycles*dt-dt,cycles)
Deleted lines 166-167:
Changed lines 168-169 from:
to:
# Process
Changed lines 172-173 from:
s.solve(
# retrieve Ca and Tmeasurements from the process
# retrieve Ca and T
to:
s.solve(disp=False)
# retrieve Ca and T measurements
# retrieve Ca and T measurements
Changed line 177 from:
to:
# Estimator
Changed lines 183-185 from:
to:
# Solve MHE
m.solve(disp=False)
m.solve(disp=False)
Changed lines 197-203 from:
to:
if i==0:
# remove '.value' overrides to use '.MEAS'
for xi in [Ca,T,Tc]:
xi.value=[None]*len(s.time)
Tc_mhe.value=[None]*len(m.time)
print('MHE: Ca (est)=' + str(Ca_mhe_store[i]) + \
# remove '.value' overrides to use '.MEAS'
for xi in [Ca,T,Tc]:
xi.value=[None]*len(s.time)
Tc_mhe.value=[None]*len(m.time)
print('MHE: Ca (est)=' + str(Ca_mhe_store[i]) + \
Changed line 205 from:
' UA (estimated)=' + str(UA_mhe_store[i]) + \
to:
' UA (est)=' + str(UA_mhe_store[i]) + \
Changed line 208 from:
#%% plot results
to:
# plot results
Changed lines 48-49 from:
# Simulation
to:
#%% Simulation
Changed line 52 from:
# 1 step of simulation, discretization matches MHE
to:
#1 step of simulation, discretization matches MHE
Changed line 55 from:
# Receive measurement from simulated control
to:
#Receive measurement from simulated control
Changed line 60 from:
# State variables to watch
to:
#State variables to watch
Changed line 64 from:
# Other parameters
to:
#other parameters
Changed line 76 from:
# Variables
to:
#Variables
Changed line 80 from:
# Rate equations
to:
#Rate equations
Changed line 83 from:
# CSTR equations
to:
#CSTR equations
Changed line 87 from:
# Options
to:
#Options
Changed lines 93-96 from:
# MHE
# Model
#
to:
#%% MHE
#Model
#Model
Changed line 99 from:
# 6 time points in horizon
to:
#6 time points in horizon
Changed line 102 from:
# Parameter to Estimate
to:
#Parameter to Estimate
Changed line 106 from:
# Upper and lower bounds for optimizer
to:
#upper and lower bounds for optimizer
Changed line 110 from:
# Measurement input
to:
#Measurement input
Changed line 115 from:
# Measurement to match simulation with
to:
#Measurement to match simulation with
Changed line 121 from:
# State to watch
to:
#State to watch
Changed line 124 from:
# Other parameters
to:
#Other parameters
Changed line 135 from:
# Equation variables (2 other DOF from CV and FV)
to:
#Equation variables(2 other DOF from CV and FV)
Changed line 139 from:
# Reaction equations
to:
#Reaction equations
Changed line 142 from:
# CSTR equations
to:
#CSTR equations
Changed lines 146-147 from:
# Global Tuning
to:
#Global Tuning
Changed lines 153-154 from:
# Loop
to:
#%% Loop
Added lines 172-173:
Changed line 211 from:
# plot results
to:
#%% plot results
Changed lines 36-236 from:
to:
# get latest gekko packge with:
# pip install gekko
# or
# pip install gekko --upgrade
# to upgrade the version to the latest
# or from the Python script:
# import pip
# pip.main(['install','gekko'])
from gekko import GEKKO
import numpy as np
import matplotlib.pyplot as plt
# Simulation
s = GEKKO(name='cstr-sim')
# 1 step of simulation, discretization matches MHE
s.time = np.linspace(0,.1,2)
# Receive measurement from simulated control
Tc = s.MV(value=300,name='tc')
Tc.FSTATUS = 1 #receive measurement
Tc.STATUS = 0 #don't optimize
# State variables to watch
Ca = s.SV(value=.7, ub=1, lb=0,name='ca')
T = s.SV(value=305,lb=250,ub=500,name='t')
# Other parameters
q = s.Param(value=100)
V = s.Param(value=100)
rho = s.Param(value=1000)
Cp = s.Param(value=0.239)
mdelH = s.Param(value=50000)
ER = s.Param(value=8750)
k0 = s.Param(value=7.2*10**10)
UA = s.Param(value=5*10**4)
Ca0 = s.Param(value=1)
T0 = s.Param(value=350)
# Variables
k = s.Var()
rate = s.Var()
# Rate equations
s.Equation(k==k0*s.exp(-ER/T))
s.Equation(rate==k*Ca)
# CSTR equations
s.Equation(V* Ca.dt() == q*(Ca0-Ca)-V*rate)
s.Equation(rho*Cp*V* T.dt() == q*rho*Cp*(T0-T) + V*mdelH*rate + UA*(Tc-T))
# Options
s.options.IMODE = 4 #dynamic simulation
s.options.NODES = 3
s.options.SOLVER = 3
# MHE
# Model
m = GEKKO(name='cstr-mhe')
# 6 time points in horizon
m.time = np.linspace(0,.5,6)
# Parameter to Estimate
UA_mhe = m.FV(value=10*10**4,name='ua')
UA_mhe.STATUS = 1 #estimate
UA_mhe.FSTATUS = 0 #no measurements
# Upper and lower bounds for optimizer
UA_mhe.LOWER = 10000
UA_mhe.UPPER = 100000
# Measurement input
Tc_mhe = m.MV(value=300,name='tc')
Tc_mhe.STATUS = 0 #don't estimate
Tc_mhe.FSTATUS = 1 #receive measurement
# Measurement to match simulation with
T_mhe = m.CV(value=325 ,lb=250,ub=500,name='t')
T_mhe.STATUS = 1 #minimize error between simulation and measurement
T_mhe.FSTATUS = 1 #receive measurement
T_mhe.MEAS_GAP = 0.1 #measurement deadband gap
# State to watch
Ca_mhe = m.SV(value=0.8, ub=1, lb=0,name='ca')
# Other parameters
q = m.Param(value=100)
V = m.Param(value=100)
rho = m.Param(value=1000)
Cp = m.Param(value=0.239)
mdelH = m.Param(value=50000)
ER = m.Param(value=8750)
k0 = m.Param(value=7.2*10**10)
Ca0 = m.Param(value=1)
T0 = m.Param(value=350)
# Equation variables (2 other DOF from CV and FV)
k = m.Var()
rate = m.Var()
# Reaction equations
m.Equation(k==k0*m.exp(-ER/T_mhe))
m.Equation(rate==k*Ca_mhe)
# CSTR equations
m.Equation(V* Ca_mhe.dt() == q*(Ca0-Ca_mhe)-V*rate) #mol balance
m.Equation(rho*Cp*V* T_mhe.dt() == q*rho*Cp*(T0-T_mhe) + V*mdelH*rate + UA_mhe*(Tc_mhe-T_mhe)) #energy balance
# Global Tuning
m.options.IMODE = 5 #MHE
m.options.EV_TYPE = 1
m.options.NODES = 3
m.options.SOLVER = 3 #IPOPT
# Loop
# number of cycles to run
cycles = 50
# step in the jacket cooling temperature at cycle 6
Tc_meas = np.empty(cycles)
Tc_meas[0:15] = 280
Tc_meas[5:cycles] = 300
dt = 0.1 # min
time = np.linspace(0,cycles*dt-dt,cycles) # time points for plot
# allocate storage
Ca_meas = np.empty(cycles)
T_meas = np.empty(cycles)
UA_mhe_store = np.empty(cycles)
Ca_mhe_store = np.empty(cycles)
T_mhe_store = np.empty(cycles)
for i in range(cycles):
## Process
# input Tc (jacket cooling temperature)
Tc.MEAS = Tc_meas[i]
# simulate process model, 1 time step
s.solve()
# retrieve Ca and T measurements from the process
Ca_meas[i] = Ca.MODEL
T_meas[i] = T.MODEL
## Estimator
# input process measurements
# input Tc (jacket cooling temperature)
Tc_mhe.MEAS = Tc_meas[i]
# input T (reactor temperature)
T_mhe.MEAS = T_meas[i] #CV
# solve process model, 1 time step
m.solve()
# check if successful
if m.options.APPSTATUS == 1:
# retrieve solution
UA_mhe_store[i] = UA_mhe.NEWVAL
Ca_mhe_store[i] = Ca_mhe.MODEL
T_mhe_store[i] = T_mhe.MODEL
else:
# failed solution
UA_mhe_store[i] = 0
Ca_mhe_store[i] = 0
T_mhe_store[i] = 0
print('MHE results: Ca (estimated)=' + str(Ca_mhe_store[i]) + \
' Ca (actual)=' + str(Ca_meas[i]) + \
' UA (estimated)=' + str(UA_mhe_store[i]) + \
' UA (actual)=50000')
# plot results
plt.figure()
plt.subplot(411)
plt.plot(time,Tc_meas,'k-',linewidth=2)
plt.axis([0,time[-1],275,305])
plt.ylabel('Jacket T (K)')
plt.legend('T_c')
plt.subplot(412)
plt.plot([0,time[-1]],[50000,50000],'k--')
plt.plot(time,UA_mhe_store,'r:',linewidth=2)
plt.axis([0,time[-1],10000,100000])
plt.ylabel('UA')
plt.legend(['Actual UA','Predicted UA'],loc=4)
plt.subplot(413)
plt.plot(time,T_meas,'ro')
plt.plot(time,T_mhe_store,'b-',linewidth=2)
plt.axis([0,time[-1],300,340])
plt.ylabel('Reactor T (K)')
plt.legend(['Measured T','Predicted T'],loc=4)
plt.subplot(414)
plt.plot(time,Ca_meas,'go')
plt.plot(time,Ca_mhe_store,'m-',linewidth=2)
plt.axis([0,time[-1],.6,1])
plt.ylabel('Reactor C_a (mol/L)')
plt.legend(['Measured C_a','Predicted C_a'],loc=4)
plt.show()
# pip install gekko
# or
# pip install gekko --upgrade
# to upgrade the version to the latest
# or from the Python script:
# import pip
# pip.main(['install','gekko'])
from gekko import GEKKO
import numpy as np
import matplotlib.pyplot as plt
# Simulation
s = GEKKO(name='cstr-sim')
# 1 step of simulation, discretization matches MHE
s.time = np.linspace(0,.1,2)
# Receive measurement from simulated control
Tc = s.MV(value=300,name='tc')
Tc.FSTATUS = 1 #receive measurement
Tc.STATUS = 0 #don't optimize
# State variables to watch
Ca = s.SV(value=.7, ub=1, lb=0,name='ca')
T = s.SV(value=305,lb=250,ub=500,name='t')
# Other parameters
q = s.Param(value=100)
V = s.Param(value=100)
rho = s.Param(value=1000)
Cp = s.Param(value=0.239)
mdelH = s.Param(value=50000)
ER = s.Param(value=8750)
k0 = s.Param(value=7.2*10**10)
UA = s.Param(value=5*10**4)
Ca0 = s.Param(value=1)
T0 = s.Param(value=350)
# Variables
k = s.Var()
rate = s.Var()
# Rate equations
s.Equation(k==k0*s.exp(-ER/T))
s.Equation(rate==k*Ca)
# CSTR equations
s.Equation(V* Ca.dt() == q*(Ca0-Ca)-V*rate)
s.Equation(rho*Cp*V* T.dt() == q*rho*Cp*(T0-T) + V*mdelH*rate + UA*(Tc-T))
# Options
s.options.IMODE = 4 #dynamic simulation
s.options.NODES = 3
s.options.SOLVER = 3
# MHE
# Model
m = GEKKO(name='cstr-mhe')
# 6 time points in horizon
m.time = np.linspace(0,.5,6)
# Parameter to Estimate
UA_mhe = m.FV(value=10*10**4,name='ua')
UA_mhe.STATUS = 1 #estimate
UA_mhe.FSTATUS = 0 #no measurements
# Upper and lower bounds for optimizer
UA_mhe.LOWER = 10000
UA_mhe.UPPER = 100000
# Measurement input
Tc_mhe = m.MV(value=300,name='tc')
Tc_mhe.STATUS = 0 #don't estimate
Tc_mhe.FSTATUS = 1 #receive measurement
# Measurement to match simulation with
T_mhe = m.CV(value=325 ,lb=250,ub=500,name='t')
T_mhe.STATUS = 1 #minimize error between simulation and measurement
T_mhe.FSTATUS = 1 #receive measurement
T_mhe.MEAS_GAP = 0.1 #measurement deadband gap
# State to watch
Ca_mhe = m.SV(value=0.8, ub=1, lb=0,name='ca')
# Other parameters
q = m.Param(value=100)
V = m.Param(value=100)
rho = m.Param(value=1000)
Cp = m.Param(value=0.239)
mdelH = m.Param(value=50000)
ER = m.Param(value=8750)
k0 = m.Param(value=7.2*10**10)
Ca0 = m.Param(value=1)
T0 = m.Param(value=350)
# Equation variables (2 other DOF from CV and FV)
k = m.Var()
rate = m.Var()
# Reaction equations
m.Equation(k==k0*m.exp(-ER/T_mhe))
m.Equation(rate==k*Ca_mhe)
# CSTR equations
m.Equation(V* Ca_mhe.dt() == q*(Ca0-Ca_mhe)-V*rate) #mol balance
m.Equation(rho*Cp*V* T_mhe.dt() == q*rho*Cp*(T0-T_mhe) + V*mdelH*rate + UA_mhe*(Tc_mhe-T_mhe)) #energy balance
# Global Tuning
m.options.IMODE = 5 #MHE
m.options.EV_TYPE = 1
m.options.NODES = 3
m.options.SOLVER = 3 #IPOPT
# Loop
# number of cycles to run
cycles = 50
# step in the jacket cooling temperature at cycle 6
Tc_meas = np.empty(cycles)
Tc_meas[0:15] = 280
Tc_meas[5:cycles] = 300
dt = 0.1 # min
time = np.linspace(0,cycles*dt-dt,cycles) # time points for plot
# allocate storage
Ca_meas = np.empty(cycles)
T_meas = np.empty(cycles)
UA_mhe_store = np.empty(cycles)
Ca_mhe_store = np.empty(cycles)
T_mhe_store = np.empty(cycles)
for i in range(cycles):
## Process
# input Tc (jacket cooling temperature)
Tc.MEAS = Tc_meas[i]
# simulate process model, 1 time step
s.solve()
# retrieve Ca and T measurements from the process
Ca_meas[i] = Ca.MODEL
T_meas[i] = T.MODEL
## Estimator
# input process measurements
# input Tc (jacket cooling temperature)
Tc_mhe.MEAS = Tc_meas[i]
# input T (reactor temperature)
T_mhe.MEAS = T_meas[i] #CV
# solve process model, 1 time step
m.solve()
# check if successful
if m.options.APPSTATUS == 1:
# retrieve solution
UA_mhe_store[i] = UA_mhe.NEWVAL
Ca_mhe_store[i] = Ca_mhe.MODEL
T_mhe_store[i] = T_mhe.MODEL
else:
# failed solution
UA_mhe_store[i] = 0
Ca_mhe_store[i] = 0
T_mhe_store[i] = 0
print('MHE results: Ca (estimated)=' + str(Ca_mhe_store[i]) + \
' Ca (actual)=' + str(Ca_meas[i]) + \
' UA (estimated)=' + str(UA_mhe_store[i]) + \
' UA (actual)=50000')
# plot results
plt.figure()
plt.subplot(411)
plt.plot(time,Tc_meas,'k-',linewidth=2)
plt.axis([0,time[-1],275,305])
plt.ylabel('Jacket T (K)')
plt.legend('T_c')
plt.subplot(412)
plt.plot([0,time[-1]],[50000,50000],'k--')
plt.plot(time,UA_mhe_store,'r:',linewidth=2)
plt.axis([0,time[-1],10000,100000])
plt.ylabel('UA')
plt.legend(['Actual UA','Predicted UA'],loc=4)
plt.subplot(413)
plt.plot(time,T_meas,'ro')
plt.plot(time,T_mhe_store,'b-',linewidth=2)
plt.axis([0,time[-1],300,340])
plt.ylabel('Reactor T (K)')
plt.legend(['Measured T','Predicted T'],loc=4)
plt.subplot(414)
plt.plot(time,Ca_meas,'go')
plt.plot(time,Ca_mhe_store,'m-',linewidth=2)
plt.axis([0,time[-1],.6,1])
plt.ylabel('Reactor C_a (mol/L)')
plt.legend(['Measured C_a','Predicted C_a'],loc=4)
plt.show()
Added lines 32-38:
(:toggle hide gekko button show="Show GEKKO (Python) Code":)
(:div id=gekko:)
(:source lang=python:)
# add source
(:sourceend:)
(:divend:)
Changed lines 11-12 from:
'''Objective:''' Design an estimator to predict an unknown parameter and state variable. Develop a model of the reactor and implement the model with moving horizon estimation in MATLAB, Python, or Simulink. ''Estimated time: 3 hours.''
to:
'''Objective:''' Design an estimator to predict an unknown parameter and state variable. Use a model of the reactor and implement the estimator to detect the current states (temperature and concentration) as well as the unmeasured heat transfer coefficient (U). ''Estimated time: 2-3 hours.''
Added lines 23-24:
The estimator can be any type such as a Kalman filter, Extended Kalman filter, Unscented Kalman Filter (particle filter), or an observer that can detect the states (T and Ca) along with the unknown parameter (U). The following solutions demonstrate an implementation of Moving Horizon Estimation.
Changed lines 31-39 from:
(:htmlend:)
to:
(:htmlend:)
!!!! References
# Haseltine, E.L., Rawlings, J.B., Critical Evaluation of Extended Kalman Filtering and Moving-Horizon Estimation, Industrial & Engineering Chemistry Research 2005 44 (8), 2451-2460, DOI: 10.1021/ie034308l [[https://jbrwww.che.wisc.edu/tech-reports/twmcc-2002-03.pdf|Preprint]], [[https://pubs.acs.org/doi/abs/10.1021/ie034308l|Article]]
# Spivey, B.J., Hedengren, J.D., Edgar, T.F., Constrained Nonlinear Estimation for Industrial Process Fouling, Industrial & Engineering Chemistry Research 2010 49 (17), 7824-7831, DOI: 10.1021/ie9018116 [[https://pubs.acs.org/doi/abs/10.1021/ie9018116|Article]]
# Hedengren, J. D., Eaton, A. N., Overview of Estimation Methods for Industrial Dynamic Systems, Optimization and Engineering, Springer, Vol 18 (1), 2017, pp. 155-178, DOI: 10.1007/s11081-015-9295-9. [[https://apm.byu.edu/prism/uploads/Members/eaton_hedengren_OPTE_springer.pdf|Preprint]], [[https://link.springer.com/article/10.1007/s11081-015-9295-9|Article]]
!!!! References
# Haseltine, E.L., Rawlings, J.B., Critical Evaluation of Extended Kalman Filtering and Moving-Horizon Estimation, Industrial & Engineering Chemistry Research 2005 44 (8), 2451-2460, DOI: 10.1021/ie034308l [[https://jbrwww.che.wisc.edu/tech-reports/twmcc-2002-03.pdf|Preprint]], [[https://pubs.acs.org/doi/abs/10.1021/ie034308l|Article]]
# Spivey, B.J., Hedengren, J.D., Edgar, T.F., Constrained Nonlinear Estimation for Industrial Process Fouling, Industrial & Engineering Chemistry Research 2010 49 (17), 7824-7831, DOI: 10.1021/ie9018116 [[https://pubs.acs.org/doi/abs/10.1021/ie9018116|Article]]
# Hedengren, J. D., Eaton, A. N., Overview of Estimation Methods for Industrial Dynamic Systems, Optimization and Engineering, Springer, Vol 18 (1), 2017, pp. 155-178, DOI: 10.1007/s11081-015-9295-9. [[https://apm.byu.edu/prism/uploads/Members/eaton_hedengren_OPTE_springer.pdf|Preprint]], [[https://link.springer.com/article/10.1007/s11081-015-9295-9|Article]]
Deleted lines 7-8:
Attach:download.png [[Attach:mhe_simulink.zip|MHE Example in Simulink]]
Added lines 8-9:
Attach:download.png [[Attach:mhe_simulink.zip|MHE Example in Simulink]]
Changed lines 25-29 from:
Attach:download.png [[Attach:cstr_mhe_solution.zip|CSTR MHE Solution in Simulink]]
to:
Attach:download.png [[Attach:cstr_mhe_solution.zip|CSTR MHE Solution in Simulink]]
(:html:)
<iframe width="560" height="315" src="https://www.youtube.com/embed/yQWgSByYjd8?rel=0" frameborder="0" allowfullscreen></iframe>
(:htmlend:)
(:html:)
<iframe width="560" height="315" src="https://www.youtube.com/embed/yQWgSByYjd8?rel=0" frameborder="0" allowfullscreen></iframe>
(:htmlend:)
Changed line 11 from:
'''Objective:''' Design an estimator to predict an unknown parameter and state variable. Develop a model of the reactor and implement the model with moving horizon estimation in Simulink. ''Estimated time: 3 hours.''
to:
'''Objective:''' Design an estimator to predict an unknown parameter and state variable. Develop a model of the reactor and implement the model with moving horizon estimation in MATLAB, Python, or Simulink. ''Estimated time: 3 hours.''
Changed lines 17-19 from:
A reactor is used to convert a hazardous chemical '''A''' to an acceptable chemical '''B''' in waste stream before entering a nearby lake. This particular reactor is dynamically modeled as a Continuously Stirred Tank Reactor (CSTR) with a simplified kinetic mechanism that describes the conversion of reactant '''A''' to product '''B''' with an irreversible and exothermic reaction. It is desired to maintain the temperature at a constant setpoint that maximizes the destruction of A (highest possible temperature).
Design an estimator to predict the concentration ofleaving the reactor and the heat transfer coefficient ''UA'' from the measured reactor temperature ''T'' and jacket temperature ''T'_c_'''. See a [[https://apmonitor.com/che436/index.php/Main/CaseStudyCSTR|related CSTR case study]] for details on the model.
Design an estimator to predict the concentration of
to:
A reactor is used to convert a hazardous chemical '''A''' to an acceptable chemical '''B''' in waste stream before entering a nearby lake. This particular reactor is dynamically modeled as a Continuously Stirred Tank Reactor (CSTR) with a simplified kinetic mechanism that describes the conversion of reactant '''A''' to product '''B''' with an irreversible and exothermic reaction. It is desired to maintain the temperature at a constant setpoint that maximizes the destruction of '''A''' (highest possible temperature). First, however, an estimator must predict the concentration of '''A''' because there is no direct measurement of this quantity. The reaction kinetics and dynamic equations are well-known but there is a parameter in the model, the heat transfer coefficient '''UA''', that is unknown.
Design an estimator to predict the concentration of '''A''' leaving the reactor and the heat transfer coefficient '''UA''' from the measured reactor temperature '''T''' and jacket temperature '''T'_c_''''. See a [[https://apmonitor.com/che436/index.php/Main/CaseStudyCSTR|related CSTR case study]] for details on the model.
Design an estimator to predict the concentration of '''A''' leaving the reactor and the heat transfer coefficient '''UA''' from the measured reactor temperature '''T''' and jacket temperature '''T'_c_''''. See a [[https://apmonitor.com/che436/index.php/Main/CaseStudyCSTR|related CSTR case study]] for details on the model.
Changed lines 23-24 from:
Attach:download.png [[Attach:cstr_mhe_solution_python.zip|CSTR MHE Solution in Python]]Attach:download.png [[Attach:cstr_mhe_solution_matlab.zip|CSTR MHE Solution in MATLAB]]
to:
Attach:download.png [[Attach:cstr_mhe_solution_matlab.zip|CSTR MHE Solution in MATLAB]]
Attach:download.png [[Attach:cstr_mhe_solution_python.zip|CSTR MHE Solution in Python]]
Attach:download.png [[Attach:cstr_mhe_solution_python.zip|CSTR MHE Solution in Python]]
Changed lines 23-24 from:
Attach:download.png [[Attach:cstr_mhe_solution.zip|CSTR MHE Solution in Simulink]]
to:
Attach:download.png [[Attach:cstr_mhe_solution_python.zip|CSTR MHE Solution in Python]]Attach:download.png [[Attach:cstr_mhe_solution_matlab.zip|CSTR MHE Solution in MATLAB]]
Attach:download.png [[Attach:cstr_mhe_solution.zip|CSTR MHE Solution in Simulink]]
Attach:download.png [[Attach:cstr_mhe_solution.zip|CSTR MHE Solution in Simulink]]
Changed line 19 from:
Design an estimator to predict the concentration of ''A'' leaving the reactor and the heat transfer coefficient ''UA'' from the measured reactor temperature ''T'' and jacket temperature ''T'_c_'''.
to:
Design an estimator to predict the concentration of ''A'' leaving the reactor and the heat transfer coefficient ''UA'' from the measured reactor temperature ''T'' and jacket temperature ''T'_c_'''. See a [[https://apmonitor.com/che436/index.php/Main/CaseStudyCSTR|related CSTR case study]] for details on the model.
Changed line 23 from:
Attach:download.png [[Attach:cstr_mhe_solution.zip|CSTR MHE Solution]]
to:
Attach:download.png [[Attach:cstr_mhe_solution.zip|CSTR MHE Solution in Simulink]]
Changed lines 21-23 from:
!!!! Solution
to:
!!!! Solution
Attach:download.png [[Attach:cstr_mhe_solution.zip|CSTR MHE Solution]]
Attach:download.png [[Attach:cstr_mhe_solution.zip|CSTR MHE Solution]]
Changed line 19 from:
Design an estimator to predict the concentration of ''A'' leaving the reactor and the heat transfer coefficient ''UA'' from the measured reactor temperature ''T'' and jacket temperature ''T'_c'''.
to:
Design an estimator to predict the concentration of ''A'' leaving the reactor and the heat transfer coefficient ''UA'' from the measured reactor temperature ''T'' and jacket temperature ''T'_c_'''.
Changed line 19 from:
Design an estimator to predict the concentration of ''A'' leaving the reactor and the heat transfer coefficient ''UA'' from the measured temperature.
to:
Design an estimator to predict the concentration of ''A'' leaving the reactor and the heat transfer coefficient ''UA'' from the measured reactor temperature ''T'' and jacket temperature ''T'_c'''.
Added lines 18-19:
Design an estimator to predict the concentration of ''A'' leaving the reactor and the heat transfer coefficient ''UA'' from the measured temperature.
Changed line 13 from:
Attach:download.png [[Attach:cstr_control.zip|CSTR Source Files]]
to:
Attach:download.png [[Attach:cstr_estimation.zip|CSTR Source Files]]
Added lines 12-13:
Attach:download.png [[Attach:cstr_control.zip|CSTR Source Files]]
Added lines 9-17:
!!!! Exercise
'''Objective:''' Design an estimator to predict an unknown parameter and state variable. Develop a model of the reactor and implement the model with moving horizon estimation in Simulink. ''Estimated time: 3 hours.''
Attach:cstr.png
A reactor is used to convert a hazardous chemical '''A''' to an acceptable chemical '''B''' in waste stream before entering a nearby lake. This particular reactor is dynamically modeled as a Continuously Stirred Tank Reactor (CSTR) with a simplified kinetic mechanism that describes the conversion of reactant '''A''' to product '''B''' with an irreversible and exothermic reaction. It is desired to maintain the temperature at a constant setpoint that maximizes the destruction of A (highest possible temperature).
!!!! Solution
'''Objective:''' Design an estimator to predict an unknown parameter and state variable. Develop a model of the reactor and implement the model with moving horizon estimation in Simulink. ''Estimated time: 3 hours.''
Attach:cstr.png
A reactor is used to convert a hazardous chemical '''A''' to an acceptable chemical '''B''' in waste stream before entering a nearby lake. This particular reactor is dynamically modeled as a Continuously Stirred Tank Reactor (CSTR) with a simplified kinetic mechanism that describes the conversion of reactant '''A''' to product '''B''' with an irreversible and exothermic reaction. It is desired to maintain the temperature at a constant setpoint that maximizes the destruction of A (highest possible temperature).
!!!! Solution
Changed lines 7-8 from:
Moving Horizon Estimation (MHE) uses dynamic optimization and a backward time horizon of measurements to optimally adjust parameters and states. The data may include noise (random fluctuations), drift (gradual departure from true values), outliers (sudden and temporary departure from true values), or other inaccuracies. Nonlinear programming solvers are employed to numerically converge the dynamic optimization problem.
to:
Moving Horizon Estimation (MHE) uses dynamic optimization and a backward time horizon of measurements to optimally adjust parameters and states. The data may include noise (random fluctuations), drift (gradual departure from true values), outliers (sudden and temporary departure from true values), or other inaccuracies. Nonlinear programming solvers are employed to numerically converge the dynamic optimization problem.
Added lines 4-5:
Dynamic models constructed with equations that describe physical phenomena may need to be tuned by adjusting parameters so that predicted outputs match with experimental data. Physical models are based on the underlying physical principles that govern the problem and result from expressions such as a force or momentum balance and may include quantities such as velocity, acceleration, and position. Other quantities of interest may include anything that changes with respect to time such as reactor composition, temperature, mole fraction, etc. Mathematical models likely contain both physical and experimental elements. This section shows how to reconcile experimental data with the physical model through parameter estimation.
Added lines 1-5:
(:title Moving Horizon Estimation:)
(:keywords Kalman filter, Simulink, moving horizon, time window, dynamic data, validation, estimation, differential, algebraic, tutorial:)
(:description Dynamic state and parameter estimation with Moving Horizon Estimation:)
Moving Horizon Estimation (MHE) uses dynamic optimization and a backward time horizon of measurements to optimally adjust parameters and states. The data may include noise (random fluctuations), drift (gradual departure from true values), outliers (sudden and temporary departure from true values), or other inaccuracies. Nonlinear programming solvers are employed to numerically converge the dynamic optimization problem.
(:keywords Kalman filter, Simulink, moving horizon, time window, dynamic data, validation, estimation, differential, algebraic, tutorial:)
(:description Dynamic state and parameter estimation with Moving Horizon Estimation:)
Moving Horizon Estimation (MHE) uses dynamic optimization and a backward time horizon of measurements to optimally adjust parameters and states. The data may include noise (random fluctuations), drift (gradual departure from true values), outliers (sudden and temporary departure from true values), or other inaccuracies. Nonlinear programming solvers are employed to numerically converge the dynamic optimization problem.