SlideShare una empresa de Scribd logo
1 de 33
Descargar para leer sin conexión
System Identification Seminar on
Kalman Filtering
By: Mahsa Rezaei (909238)
School of Electrical and Computer Engineering
February 2012
1
Outline:
 Motivation
 Historical Prospective
 What is a Kalman filter?
 Applications
 Advantages
 The problem
 How does Kalman filter work?
 Estimation in linear systems
 What criteria should our estimator satisfy?
 Preparation before going to Algorithm
 Standard Kalman filter Algorithm
 Example on Kalman filter for linear systems (Vehicle Navigation)
 Shift to nonlinear systems
 Extended Kalman filter Algorithm
 Example on Extended Kalman filter for Nonlinear Systems
 References 2
Motivation:
filtering is desirable in many situations in engineering and
embedded systems.
For example, radio communication signals are corrupted with
noise.
A good filtering algorithm can remove the noise from
electromagnetic signals while retaining the useful information.
Another example is power supply voltages. Uninterruptible power
supplies are devices that filter line voltages in order to smooth out
undesirable fluctuations that might otherwise shorten the lifespan
of electrical devices such as computers and printers.
3
Historical Prospective:
In 1960 and 1961 Rudolf Emil Kalman published his papers
on a recursive predictive filter that is based on the use of state
space techniques and recursive algorithms and therewith he
revolutionized the field of estimation.
Like many new technologies, the Kalman filter was developed
to solve a specific problem, in this case, spacecraft navigation
for the Apollo space program. Since that time the so-called
Kalman filter has been the subject of extensive research and
applications.
Sometimes the filter is referred to as the Kalman-Bucy filter
because of Richard Bucy‟s early work on the topic, conducted
jointly with Kalman.
4
What is a Kalman filter?
The Kalman filter is a set of mathematical equations that
provides an efficient computational (recursive) means to
estimate the state of a dynamic system, in a way that
minimizes the mean of the squared error.
The filter is very powerful in the sense that it supports
estimations of past, present, and even future states, and it can
do so even when the precise nature of the modeled system is
unknown.
5
Applications:
The applications of a Kalman filter are numerous:
 Tracking objects (e.g., missiles, faces, heads, hands)
 All forms of navigation (aerospace, land, and marine),
 Economics
 Fuzzy logic
 Neural network training
 The detection of underground radioactivity,
 Many computer vision applications
 Many more
The principal uses of Kalman filtering have been in „Modern‟
control systems, in the tracking and navigation of all sorts of
vehicles, and predictive design of estimation and control systems.
6
Advantages:
1.The Kalman filter is implementable in the form of an algorithm
for a digital computer. This implementation may be slower, but it
is capable of much greater accuracy than had been achieved with
analog filters.
2.The Kalman filter does not require that the deterministic
dynamics or the random processes have stationary properties,
and many applications of importance include non-stationary
stochastic processes.
3.The Kalman filter is compatible with the state-space
formulation of optimal controllers for dynamic systems .
4.The Kalman filter requires less additional mathematical
preparation to learn and use than other filters.
7
The problem:
Measuring
Devices Estimator
Measurement
Error Sources
System State
(desired but not
known)
External
Controls
Observed
Measurements
Optimal
Estimate of
System State
System
Error Sources
System
Black
Box
• System state cannot be measured directly
• Need to estimate “optimally” from measurements
8
How does Kalman filter work?
The Kalman filter estimates the state of a dynamic system.
This dynamic system can be disturbed by some noise, mostly
assumed as white noise. To improve the estimated state, the
Kalman filter uses measurements that are related to the state
but disturbed as well.
Thus the Kalman filter consists of two steps:
1.The prediction
2.The correction
9
In the first step the state is predicted with the dynamic model.
In the second step it is corrected with the observation model, so that
the error covariance of the estimator is minimized. In this sense it is an
optimal estimator.
This procedure is repeated for each time step with the state of the
previous time step as initial value. Therefore the kalman filter is called
a recursive filter.
10
Estimation in linear systems:
Many physical processes, such as a vehicle driving along a road, a satellite
orbiting the earth, a motor shaft driven by winding currents, or a sinusoidal radio-
frequency carrier signal, can be approximated as linear systems.
A linear system is simply a process that can be described by the following two
equations:
𝒙 𝒌 = 𝑨 𝒌−𝟏 𝒙 𝒌−𝟏 + 𝑩 𝒌−𝟏 𝒖 𝒌−𝟏 + 𝑳 𝒌−𝟏 𝒘 𝒌−𝟏
𝒚 𝒌 = 𝑯 𝒌 𝒙 𝒌 + 𝑴 𝒌 𝒗 𝒌
𝒙 𝒏×𝟏: State vector
𝒖 𝒓×𝟏: input vector
𝒘 𝒑×𝟏: Process(state) noise
𝒚 𝒎×𝟏: measurement(output) vector
𝒗 𝒈×𝟏: measurement error
𝒌 = 𝟎, 𝟏, 𝟐, … :Discrete time index
𝑨 𝒏×𝒏: transition matrix which relates the state at time k to the one at k-1
𝑩 𝒏×𝒓: relates the optional control input u to the state x
𝑯 𝒎×𝒏: relates the measurement to the state
11
What criteria should our estimator satisfy?
We want an estimator that gives an accurate estimate of the true
state even though we cannot directly measure it. What criteria
should our estimator satisfy? Two obvious requirements come to
mind.
First, we want the average value of our state estimate to be equal
to the average value of the true state. That is, we don‟t want our
estimate to be biased one way or another.
Second, we want a state estimate that varies from the true state as
little as possible. That is, not only do we want the average of the
state estimate to be equal to the average of the true state, but we
also want an estimator that results in the smallest possible
variation of the state estimate.
12
Preparation before going to algorithm:
1.Definition:
𝒙−: a priori state, the predicted value before the update
𝒙+: a posteriori state, the corrected value after the update
2.Expected value of a random variable x is:
𝑬 𝒙 = 𝒙 𝑷 𝒙 𝒅𝒙 ≈ 𝒙 =
𝟏
𝑵
𝒙𝒊
𝑵
𝒊=𝟏
Expected value of a vector X is by component:
𝑬 𝑿 = 𝑿 = 𝑿 𝟏, 𝑿 𝟐, … , 𝑿 𝒏
𝑻
13
3.For the expressed criteria to be satisfied, certain condition
about the noise- that affects our system- should be assumed:
𝑬 𝒘 𝒌 = 𝟎
𝑬 𝒗 𝒌 = 𝟎
𝑬 𝒘 𝒌 𝒗 𝒌 = 𝟎 𝒘 𝒌~(𝟎, 𝑸 𝒌)
𝑬 𝒘 𝒌 𝒘𝒋
𝑻 = 𝟎 𝒇𝒐𝒓 𝒌 ≠ 𝒋
𝑬 𝒗 𝒌 𝒗𝒋
𝑻
= 𝟎 𝒇𝒐𝒓 𝒌 ≠ 𝒋 𝒗 𝒌~(𝟎, 𝑹 𝒌)
4.Definition:
𝑸 𝒌= 𝑬 𝒘 𝒌 𝒘 𝒌
𝑻 : Covariance of w
𝑹 𝒌 = 𝑬(𝒗 𝒌 𝒗 𝒌
𝑻) : Covariance of v
14
Standard Kalman filter algorithm:
Step 0: Initialization of the filter at k=0.
𝒙 𝟎
+
= 𝑬 𝒙 𝟎
𝑷 𝟎
+
= 𝑬 𝒙 𝟎 − 𝒙 𝟎
+
𝒙 𝟎 − 𝒙 𝟎
+ 𝑻
Step 1: for k=1,2,… follow the following procedure.
a)
𝒙 𝒌
−
= 𝑨 𝒌−𝟏 𝒙 𝒌−𝟏
+
+ 𝑩 𝒌−𝟏 𝒖 𝒌−𝟏
𝑷 𝒌
−
= 𝑨 𝒌−𝟏 𝑷 𝒌−𝟏
+
𝑨 𝒌−𝟏
𝑻
+ 𝑳 𝒌−𝟏 𝑸 𝒌−𝟏 𝑳 𝒌−𝟏
𝑻
b)
𝑲 𝒌 = 𝑷 𝒌
−
𝑯 𝒌
𝑻
𝑯 𝒌 𝑷 𝒌
−
𝑯 𝒌
𝑻
+ 𝑴 𝒌 𝑹 𝒌 𝑴 𝒌
𝑻 −𝟏
c)
𝒙 𝒌
+
= 𝒙 𝒌
−
+ 𝑲 𝒌 𝒚 𝒌 − 𝑯 𝒌 𝒙 𝒌
−
𝑷 𝒌
+
= (𝑰 − 𝑲 𝒌 𝑯 𝒌)𝑷 𝒌
−
15
Example on standard Kalman filter for linear
systems (Vehicle Navigation):
we want to model a vehicle going in a straight line. We can say that the
state consists of the vehicle position p and velocity v. The input u is the
commanded acceleration and the output y is the measured position. we
are able to change the acceleration and measure the position every T
seconds.
In this case, elementary laws of physics say that the position p and
velocity v will be governed by the following equations:
𝒑 𝒌 = 𝒑 𝒌−𝟏 + 𝑻𝒗 𝒌−𝟏 +
𝟏
𝟐
𝑻 𝟐
𝒖 𝒌−𝟏 + 𝒑 𝒌−𝟏
𝒗 𝒌 = 𝒗 𝒌−𝟏 + 𝑻𝒖 𝒌−𝟏 + 𝒗 𝒌
Where 𝑝 𝑘 and 𝑣 𝑘 are position and velocity noise respectively.
Now we can define a state vector x that consists of position and velocity:
𝒙 𝒌 =
𝒑 𝒌
𝒗 𝒌
16
Finally, knowing that the measured output is equal to the
position, we can write our linear system equations as follows:
𝒙 𝒌 =
𝟏 𝑻
𝟎 𝟏
𝒙 𝒌−𝟏 +
𝑻 𝟐
𝟐
𝑻
𝒖 𝒌 + 𝒘 𝒌
𝒚 𝒌 = 𝟏 𝟎 𝒙 𝒌 + 𝒛 𝒌
𝑧 𝑘 is the measurement noise due to such things as
instrumentation errors.
If we want to control the vehicle with some sort of feedback
system, we need an accurate estimate of the position p and the
velocity v. In other words, we need a way to estimate the state x.
This is where the Kalman filter comes in.
17
The position is measured with an error of 10 feet (one standard deviation).
The commanded acceleration is a constant 1 foot/sec2. The acceleration
noise is 0.2 feet/sec2 (one standard deviation). The position is measured 10
times per second (T = 0.1).
Since T = 0.1, the linear model that represents our system can be derived
from the system model presented earlier as follows:
𝒙 𝒌 =
𝟏 𝟎. 𝟏
𝟎 𝟏
𝒙 𝒌−𝟏 +
𝟎. 𝟎𝟎𝟓
𝟎. 𝟏
𝒖 𝒌 + 𝒘 𝒌
𝒚 𝒌 = 𝟏 𝟎 𝒙 𝒌 + 𝒛 𝒌
Because the standard deviation of the measurement noise is 10 feet, the 𝑅 𝑘
matrix is simply equal to 100.
Now we need to derive the 𝑄 𝑘 matrix.
18
Since the position is proportional to 0.005 times the acceleration, and the
acceleration noise is 0.2 feet/sec2, the variance of the position noise is
(0.005)2× (0.2)2= 10−6.
Similarly, since the velocity is proportional to 0.1 times the acceleration, the
variance of the velocity noise is (0.1)2× (0.2)2= 4 × 10−4.
Finally, the covariance of the position noise and velocity noise is equal to
the standard deviation of the position noise times the standard deviation of
the velocity noise, which can be calculated as 0.005 × 0.2 × 0.1 × 0.2
= 2 × 10−5.
We can combine all of these calculations to obtain the following matrix for
𝑄 𝑘 :
𝑸 𝒌 = 𝑬 𝒙𝒙 𝑻
= 𝑬
𝒑
𝒗
𝒑 𝒗 = 𝑬
𝒑 𝟐 𝒑𝒗
𝒑𝒗 𝒗 𝟐
= 𝟏𝟎−𝟔 𝟐 × 𝟏𝟎−𝟓
𝟐 × 𝟏𝟎−𝟓 𝟒 × 𝟏𝟎−𝟒
I simulated the Kalman filter for this problem using Matlab. The results are
shown in the accompanying figures.
19
Figure 1 shows the true position of the vehicle, the measured position, and
the estimated position.
20
The two smooth curves are the true position and the estimated position, and
they are almost too close to distinguish from one another. The noisy-looking
curve is the measured position. For better comparison, part of the figure is
magnified here:
21
Figure 2 shows the error between the true position and the measured
position, and the error between the true position and the Kalman filter‟s
estimated position.
22
Figure 3 shows the True and Estimated Velocity
23
Figure 4 shows the error between the true velocity and the Kalman
filter‟s estimated velocity.
24
Shift to Nonlinear Systems:
We have discussed state estimation for linear systems. But what if we want
to estimate the states of a nonlinear system?
As a matter of fact, almost all real engineering processes are nonlinear.
Some can be approximated by linear systems but some cannot.
This was recognized early in the history of Kalman filters and led to the
development of the “extended Kalman filter” which is simply an
extension of linear Kalman filter theory to nonlinear systems.
A Nonlinear system is simply a process that can be described by the
following two equations:
𝒙 𝒌 = 𝒇 𝒌−𝟏 𝒙 𝒌−𝟏, 𝒖 𝒌−𝟏 , 𝒘 𝒌−𝟏
𝒚 𝒌 = 𝒉 𝒌 𝒙 𝒌, 𝒖 𝒌 , 𝒗 𝒌
f : System function
h : Output function
25
Extended Kalman filter Algorithm:
Step 0: Initialization of the filter at k=0
𝒙 𝟎
+
= 𝑬 𝒙 𝟎
𝑷 𝟎
+
= 𝑬 𝒙 𝟎 − 𝒙 𝟎
+
𝒙 𝟎 − 𝒙 𝟎
+ 𝑻
Step 1: for k=1,2,… follow the following procedure.
a)
𝑭 𝒌−𝟏 =
𝝏𝒇 𝒌−𝟏
𝝏𝒙
𝒙 𝒌−𝟏
+
𝑳 𝒌−𝟏 =
𝝏𝒇 𝒌−𝟏
𝝏𝒘
𝒙 𝒌−𝟏
+
b)
𝒙 𝒌
−
= 𝒇 𝒌−𝟏 𝒙 𝒌−𝟏
+
, 𝒖 𝒌−𝟏, 𝟎
26
27
𝑷 𝒌
−
= 𝑭 𝒌−𝟏 𝑷 𝒌−𝟏
+
𝑭 𝒌−𝟏
𝑻
+ 𝑳 𝒌−𝟏 𝑸 𝒌−𝟏 𝑳 𝒌−𝟏
𝑻
c)
𝑯 𝒌 =
𝝏𝒉 𝒌
𝝏𝒙
𝒙 𝒌
−
𝑴 𝒌 =
𝝏𝒉 𝒌
𝝏𝒗
𝒙 𝒌
−
d)
𝑲 𝒌 = 𝑷 𝒌
−
𝑯 𝒌
𝑻
𝑯 𝒌 𝑷 𝒌
−
𝑯 𝒌
𝑻
+ 𝑴 𝒌 𝑹 𝒌 𝑴 𝒌
𝑻 −𝟏
e)
𝒙 𝒌
+
= 𝒙 𝒌
−
+ 𝑲 𝒌 𝒚 𝒌 − 𝒉 𝒌 𝒙 𝒌
−
, 𝟎
𝑷 𝒌
+
= (𝑰 − 𝑲 𝒌 𝑯 𝒌)𝑷 𝒌
−
28
Example on Extended Kalman filter for Nonlinear
systems:
Consider a second order system as below:
𝒚
𝒖
=
𝝎 𝒏
𝟐
𝒔 𝟐 + 𝟐𝜻𝝎 𝒏 𝒔 + 𝝎 𝒏
𝟐
Define 𝑏 = 2𝜁𝜔 𝑛
We want to estimate 𝝎 𝒏
𝟐.
In terms of differential equations:
𝒚 + 𝒃𝒚 + 𝝎 𝒏
𝟐
𝒚 = 𝝎 𝒏
𝟐
𝒖
And so we define the states as:
𝒙 𝟏 = 𝒚 𝒙 𝟏 = 𝒙 𝟐
𝒙 𝟐 = 𝒚 𝒙 𝟐 = −𝒃𝒙 𝟐 − 𝒙 𝟏 𝒙 𝟑 + 𝒙 𝟑 𝒖
𝒙 𝟑 = 𝝎 𝒏
𝟐 𝒙 𝟑 = 𝟎
29
Assume that:
𝝎 𝒏 = 𝟏 : true value
𝜻 = 𝟎. 𝟏 : zeta
𝑸 =
𝟏𝟎𝟎𝟎 𝟎
𝟎 𝟎. 𝟏
: Covariance of process noise
𝑹 =
𝟏𝟎 𝟎
𝟎 𝟏𝟎
: Covariance of measurement noise
The simulation results are enclosed in the accompanying figures.
30
This figure shows the true and estimated value of 𝜔 𝑛
2
31
Variance of estimation error is implemented in this figure
32
References:
1. J. S. Meditch. Stochastic optimal linear estimation and
control
2. Simon J. Julier and Jeffrey K. Uhlmann. A new extension of
the Kalman filter to nonlinear systems.
3. Mohinder S. Grewal and Angus P. Andrews. Kalman
filtering: Theory and Practice using Matlab
4. Rachel Kleinbauer . Kalman filtering Implementation with
Matlab
5. Greg Welch and Gary Bishop. An introduction to the Kalman
filter
33
Thank You

Más contenido relacionado

La actualidad más candente

A KALMAN FILTERING TUTORIAL FOR UNDERGRADUATE STUDENTS
A KALMAN FILTERING TUTORIAL FOR UNDERGRADUATE STUDENTSA KALMAN FILTERING TUTORIAL FOR UNDERGRADUATE STUDENTS
A KALMAN FILTERING TUTORIAL FOR UNDERGRADUATE STUDENTSIJCSES Journal
 
Kalman filter - Applications in Image processing
Kalman filter - Applications in Image processingKalman filter - Applications in Image processing
Kalman filter - Applications in Image processingRavi Teja
 
The extended kalman filter
The extended kalman filterThe extended kalman filter
The extended kalman filterMudit Parnami
 
Report kalman filtering
Report kalman filteringReport kalman filtering
Report kalman filteringIrfan Anjum
 
Kalman filtering and it's applications
Kalman filtering and it's applicationsKalman filtering and it's applications
Kalman filtering and it's applicationsMADHAVASAIYENDUVA
 
Kalman Filtering
Kalman FilteringKalman Filtering
Kalman FilteringEngin Gul
 
Artificial Neural Networks Lect5: Multi-Layer Perceptron & Backpropagation
Artificial Neural Networks Lect5: Multi-Layer Perceptron & BackpropagationArtificial Neural Networks Lect5: Multi-Layer Perceptron & Backpropagation
Artificial Neural Networks Lect5: Multi-Layer Perceptron & BackpropagationMohammed Bennamoun
 
Lecture Notes on Adaptive Signal Processing-1.pdf
Lecture Notes on Adaptive Signal Processing-1.pdfLecture Notes on Adaptive Signal Processing-1.pdf
Lecture Notes on Adaptive Signal Processing-1.pdfVishalPusadkar1
 
Fun and Easy Kalman filter Tutorial - Using Pokemon Example
Fun and Easy Kalman filter Tutorial - Using Pokemon ExampleFun and Easy Kalman filter Tutorial - Using Pokemon Example
Fun and Easy Kalman filter Tutorial - Using Pokemon ExampleRitesh Kanjee
 
Real time implementation of unscented kalman filter for target tracking
Real time implementation of unscented kalman filter for target trackingReal time implementation of unscented kalman filter for target tracking
Real time implementation of unscented kalman filter for target trackingIAEME Publication
 
Application of adaptive linear equalizer
Application of adaptive linear equalizerApplication of adaptive linear equalizer
Application of adaptive linear equalizerSayahnarahul
 

La actualidad más candente (20)

Kalman filter
Kalman filterKalman filter
Kalman filter
 
Kalman Filter Presentation
Kalman Filter PresentationKalman Filter Presentation
Kalman Filter Presentation
 
A KALMAN FILTERING TUTORIAL FOR UNDERGRADUATE STUDENTS
A KALMAN FILTERING TUTORIAL FOR UNDERGRADUATE STUDENTSA KALMAN FILTERING TUTORIAL FOR UNDERGRADUATE STUDENTS
A KALMAN FILTERING TUTORIAL FOR UNDERGRADUATE STUDENTS
 
Kalman filter - Applications in Image processing
Kalman filter - Applications in Image processingKalman filter - Applications in Image processing
Kalman filter - Applications in Image processing
 
Av 738 - Adaptive Filtering - Kalman Filters
Av 738 - Adaptive Filtering - Kalman Filters Av 738 - Adaptive Filtering - Kalman Filters
Av 738 - Adaptive Filtering - Kalman Filters
 
The extended kalman filter
The extended kalman filterThe extended kalman filter
The extended kalman filter
 
Kalman Filter | Statistics
Kalman Filter | StatisticsKalman Filter | Statistics
Kalman Filter | Statistics
 
Report kalman filtering
Report kalman filteringReport kalman filtering
Report kalman filtering
 
Kalman filtering and it's applications
Kalman filtering and it's applicationsKalman filtering and it's applications
Kalman filtering and it's applications
 
Kalman Filter Basic
Kalman Filter BasicKalman Filter Basic
Kalman Filter Basic
 
Kalman Filtering
Kalman FilteringKalman Filtering
Kalman Filtering
 
Kalman Equations
Kalman EquationsKalman Equations
Kalman Equations
 
Artificial Neural Networks Lect5: Multi-Layer Perceptron & Backpropagation
Artificial Neural Networks Lect5: Multi-Layer Perceptron & BackpropagationArtificial Neural Networks Lect5: Multi-Layer Perceptron & Backpropagation
Artificial Neural Networks Lect5: Multi-Layer Perceptron & Backpropagation
 
Data fusion with kalman filtering
Data fusion with kalman filteringData fusion with kalman filtering
Data fusion with kalman filtering
 
Lecture Notes on Adaptive Signal Processing-1.pdf
Lecture Notes on Adaptive Signal Processing-1.pdfLecture Notes on Adaptive Signal Processing-1.pdf
Lecture Notes on Adaptive Signal Processing-1.pdf
 
Kalmanfilter
Kalmanfilter Kalmanfilter
Kalmanfilter
 
Fun and Easy Kalman filter Tutorial - Using Pokemon Example
Fun and Easy Kalman filter Tutorial - Using Pokemon ExampleFun and Easy Kalman filter Tutorial - Using Pokemon Example
Fun and Easy Kalman filter Tutorial - Using Pokemon Example
 
Kalman filter demonstration
Kalman filter demonstrationKalman filter demonstration
Kalman filter demonstration
 
Real time implementation of unscented kalman filter for target tracking
Real time implementation of unscented kalman filter for target trackingReal time implementation of unscented kalman filter for target tracking
Real time implementation of unscented kalman filter for target tracking
 
Application of adaptive linear equalizer
Application of adaptive linear equalizerApplication of adaptive linear equalizer
Application of adaptive linear equalizer
 

Similar a Kalman_filtering

Inverted Pendulum Control System
Inverted Pendulum Control SystemInverted Pendulum Control System
Inverted Pendulum Control SystemAniket Govindaraju
 
Understanding kalman filter for soc estimation.
Understanding kalman filter for soc estimation.Understanding kalman filter for soc estimation.
Understanding kalman filter for soc estimation.Ratul
 
Vibration isolation progect lego(r)
Vibration isolation progect lego(r)Vibration isolation progect lego(r)
Vibration isolation progect lego(r)Open Adaptronik
 
Linear control system Open loop & Close loop Systems
Linear control system Open loop & Close loop SystemsLinear control system Open loop & Close loop Systems
Linear control system Open loop & Close loop SystemsSohaibUllah5
 
presentation.ppt
presentation.pptpresentation.ppt
presentation.pptWasiqAli28
 
07 image filtering of colored noise based on kalman filter
07 image filtering of colored noise based on kalman filter07 image filtering of colored noise based on kalman filter
07 image filtering of colored noise based on kalman filterstudymate
 
Refining Underwater Target Localization and Tracking Estimates
Refining Underwater Target Localization and Tracking EstimatesRefining Underwater Target Localization and Tracking Estimates
Refining Underwater Target Localization and Tracking EstimatesCSCJournals
 
lecture_18-19_state_observer_design.pptx
lecture_18-19_state_observer_design.pptxlecture_18-19_state_observer_design.pptx
lecture_18-19_state_observer_design.pptxAnshulShekhar3
 
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTER
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTERA NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTER
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTERijcseit
 
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTER
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTERA NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTER
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTERijcseit
 
Sensor Fusion Study - Ch5. The discrete-time Kalman filter [박정은]
Sensor Fusion Study - Ch5. The discrete-time Kalman filter  [박정은]Sensor Fusion Study - Ch5. The discrete-time Kalman filter  [박정은]
Sensor Fusion Study - Ch5. The discrete-time Kalman filter [박정은]AI Robotics KR
 
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTER
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTERA NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTER
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTERijcseit
 
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTER
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTERA NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTER
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTERijcseit
 
The describing function
The describing functionThe describing function
The describing functionkatamthreveni
 
Controls Based Q Measurement Report
Controls Based Q Measurement ReportControls Based Q Measurement Report
Controls Based Q Measurement ReportLouis Gitelman
 

Similar a Kalman_filtering (20)

Av 738-Adaptive Filters - Extended Kalman Filter
Av 738-Adaptive Filters - Extended Kalman FilterAv 738-Adaptive Filters - Extended Kalman Filter
Av 738-Adaptive Filters - Extended Kalman Filter
 
Kalman filter.pdf
Kalman filter.pdfKalman filter.pdf
Kalman filter.pdf
 
Basics Of Kalman Filter And Position Estimation Of Front Wheel Automatic Stee...
Basics Of Kalman Filter And Position Estimation Of Front Wheel Automatic Stee...Basics Of Kalman Filter And Position Estimation Of Front Wheel Automatic Stee...
Basics Of Kalman Filter And Position Estimation Of Front Wheel Automatic Stee...
 
Inverted Pendulum Control System
Inverted Pendulum Control SystemInverted Pendulum Control System
Inverted Pendulum Control System
 
Understanding kalman filter for soc estimation.
Understanding kalman filter for soc estimation.Understanding kalman filter for soc estimation.
Understanding kalman filter for soc estimation.
 
Vibration isolation progect lego(r)
Vibration isolation progect lego(r)Vibration isolation progect lego(r)
Vibration isolation progect lego(r)
 
Linear control system Open loop & Close loop Systems
Linear control system Open loop & Close loop SystemsLinear control system Open loop & Close loop Systems
Linear control system Open loop & Close loop Systems
 
presentation.ppt
presentation.pptpresentation.ppt
presentation.ppt
 
07 image filtering of colored noise based on kalman filter
07 image filtering of colored noise based on kalman filter07 image filtering of colored noise based on kalman filter
07 image filtering of colored noise based on kalman filter
 
Refining Underwater Target Localization and Tracking Estimates
Refining Underwater Target Localization and Tracking EstimatesRefining Underwater Target Localization and Tracking Estimates
Refining Underwater Target Localization and Tracking Estimates
 
solver (1)
solver (1)solver (1)
solver (1)
 
lecture_18-19_state_observer_design.pptx
lecture_18-19_state_observer_design.pptxlecture_18-19_state_observer_design.pptx
lecture_18-19_state_observer_design.pptx
 
Kalmanfilter
KalmanfilterKalmanfilter
Kalmanfilter
 
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTER
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTERA NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTER
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTER
 
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTER
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTERA NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTER
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTER
 
Sensor Fusion Study - Ch5. The discrete-time Kalman filter [박정은]
Sensor Fusion Study - Ch5. The discrete-time Kalman filter  [박정은]Sensor Fusion Study - Ch5. The discrete-time Kalman filter  [박정은]
Sensor Fusion Study - Ch5. The discrete-time Kalman filter [박정은]
 
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTER
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTERA NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTER
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTER
 
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTER
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTERA NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTER
A NEW METHOD OF SMALL-SIGNAL CALIBRATION BASED ON KALMAN FILTER
 
The describing function
The describing functionThe describing function
The describing function
 
Controls Based Q Measurement Report
Controls Based Q Measurement ReportControls Based Q Measurement Report
Controls Based Q Measurement Report
 

Más de mahsa rezaei

Excel Solver(By Mahsa Rezaei)
Excel Solver(By Mahsa Rezaei)Excel Solver(By Mahsa Rezaei)
Excel Solver(By Mahsa Rezaei)mahsa rezaei
 
seminar_presentation
seminar_presentationseminar_presentation
seminar_presentationmahsa rezaei
 
Master of Science Thesis in Electrical Engineering
Master of Science Thesis in Electrical EngineeringMaster of Science Thesis in Electrical Engineering
Master of Science Thesis in Electrical Engineeringmahsa rezaei
 
bachelor final project
bachelor final projectbachelor final project
bachelor final projectmahsa rezaei
 

Más de mahsa rezaei (9)

Excel Solver(By Mahsa Rezaei)
Excel Solver(By Mahsa Rezaei)Excel Solver(By Mahsa Rezaei)
Excel Solver(By Mahsa Rezaei)
 
seminar_presentation
seminar_presentationseminar_presentation
seminar_presentation
 
Master of Science Thesis in Electrical Engineering
Master of Science Thesis in Electrical EngineeringMaster of Science Thesis in Electrical Engineering
Master of Science Thesis in Electrical Engineering
 
thesis1
thesis1thesis1
thesis1
 
bachelor final project
bachelor final projectbachelor final project
bachelor final project
 
BPMN and Bizagi
BPMN and BizagiBPMN and Bizagi
BPMN and Bizagi
 
data mining
data miningdata mining
data mining
 
BIandDataMining
BIandDataMiningBIandDataMining
BIandDataMining
 
Mahsa_Rezaei_BPMS
Mahsa_Rezaei_BPMSMahsa_Rezaei_BPMS
Mahsa_Rezaei_BPMS
 

Kalman_filtering

  • 1. System Identification Seminar on Kalman Filtering By: Mahsa Rezaei (909238) School of Electrical and Computer Engineering February 2012 1
  • 2. Outline:  Motivation  Historical Prospective  What is a Kalman filter?  Applications  Advantages  The problem  How does Kalman filter work?  Estimation in linear systems  What criteria should our estimator satisfy?  Preparation before going to Algorithm  Standard Kalman filter Algorithm  Example on Kalman filter for linear systems (Vehicle Navigation)  Shift to nonlinear systems  Extended Kalman filter Algorithm  Example on Extended Kalman filter for Nonlinear Systems  References 2
  • 3. Motivation: filtering is desirable in many situations in engineering and embedded systems. For example, radio communication signals are corrupted with noise. A good filtering algorithm can remove the noise from electromagnetic signals while retaining the useful information. Another example is power supply voltages. Uninterruptible power supplies are devices that filter line voltages in order to smooth out undesirable fluctuations that might otherwise shorten the lifespan of electrical devices such as computers and printers. 3
  • 4. Historical Prospective: In 1960 and 1961 Rudolf Emil Kalman published his papers on a recursive predictive filter that is based on the use of state space techniques and recursive algorithms and therewith he revolutionized the field of estimation. Like many new technologies, the Kalman filter was developed to solve a specific problem, in this case, spacecraft navigation for the Apollo space program. Since that time the so-called Kalman filter has been the subject of extensive research and applications. Sometimes the filter is referred to as the Kalman-Bucy filter because of Richard Bucy‟s early work on the topic, conducted jointly with Kalman. 4
  • 5. What is a Kalman filter? The Kalman filter is a set of mathematical equations that provides an efficient computational (recursive) means to estimate the state of a dynamic system, in a way that minimizes the mean of the squared error. The filter is very powerful in the sense that it supports estimations of past, present, and even future states, and it can do so even when the precise nature of the modeled system is unknown. 5
  • 6. Applications: The applications of a Kalman filter are numerous:  Tracking objects (e.g., missiles, faces, heads, hands)  All forms of navigation (aerospace, land, and marine),  Economics  Fuzzy logic  Neural network training  The detection of underground radioactivity,  Many computer vision applications  Many more The principal uses of Kalman filtering have been in „Modern‟ control systems, in the tracking and navigation of all sorts of vehicles, and predictive design of estimation and control systems. 6
  • 7. Advantages: 1.The Kalman filter is implementable in the form of an algorithm for a digital computer. This implementation may be slower, but it is capable of much greater accuracy than had been achieved with analog filters. 2.The Kalman filter does not require that the deterministic dynamics or the random processes have stationary properties, and many applications of importance include non-stationary stochastic processes. 3.The Kalman filter is compatible with the state-space formulation of optimal controllers for dynamic systems . 4.The Kalman filter requires less additional mathematical preparation to learn and use than other filters. 7
  • 8. The problem: Measuring Devices Estimator Measurement Error Sources System State (desired but not known) External Controls Observed Measurements Optimal Estimate of System State System Error Sources System Black Box • System state cannot be measured directly • Need to estimate “optimally” from measurements 8
  • 9. How does Kalman filter work? The Kalman filter estimates the state of a dynamic system. This dynamic system can be disturbed by some noise, mostly assumed as white noise. To improve the estimated state, the Kalman filter uses measurements that are related to the state but disturbed as well. Thus the Kalman filter consists of two steps: 1.The prediction 2.The correction 9
  • 10. In the first step the state is predicted with the dynamic model. In the second step it is corrected with the observation model, so that the error covariance of the estimator is minimized. In this sense it is an optimal estimator. This procedure is repeated for each time step with the state of the previous time step as initial value. Therefore the kalman filter is called a recursive filter. 10
  • 11. Estimation in linear systems: Many physical processes, such as a vehicle driving along a road, a satellite orbiting the earth, a motor shaft driven by winding currents, or a sinusoidal radio- frequency carrier signal, can be approximated as linear systems. A linear system is simply a process that can be described by the following two equations: 𝒙 𝒌 = 𝑨 𝒌−𝟏 𝒙 𝒌−𝟏 + 𝑩 𝒌−𝟏 𝒖 𝒌−𝟏 + 𝑳 𝒌−𝟏 𝒘 𝒌−𝟏 𝒚 𝒌 = 𝑯 𝒌 𝒙 𝒌 + 𝑴 𝒌 𝒗 𝒌 𝒙 𝒏×𝟏: State vector 𝒖 𝒓×𝟏: input vector 𝒘 𝒑×𝟏: Process(state) noise 𝒚 𝒎×𝟏: measurement(output) vector 𝒗 𝒈×𝟏: measurement error 𝒌 = 𝟎, 𝟏, 𝟐, … :Discrete time index 𝑨 𝒏×𝒏: transition matrix which relates the state at time k to the one at k-1 𝑩 𝒏×𝒓: relates the optional control input u to the state x 𝑯 𝒎×𝒏: relates the measurement to the state 11
  • 12. What criteria should our estimator satisfy? We want an estimator that gives an accurate estimate of the true state even though we cannot directly measure it. What criteria should our estimator satisfy? Two obvious requirements come to mind. First, we want the average value of our state estimate to be equal to the average value of the true state. That is, we don‟t want our estimate to be biased one way or another. Second, we want a state estimate that varies from the true state as little as possible. That is, not only do we want the average of the state estimate to be equal to the average of the true state, but we also want an estimator that results in the smallest possible variation of the state estimate. 12
  • 13. Preparation before going to algorithm: 1.Definition: 𝒙−: a priori state, the predicted value before the update 𝒙+: a posteriori state, the corrected value after the update 2.Expected value of a random variable x is: 𝑬 𝒙 = 𝒙 𝑷 𝒙 𝒅𝒙 ≈ 𝒙 = 𝟏 𝑵 𝒙𝒊 𝑵 𝒊=𝟏 Expected value of a vector X is by component: 𝑬 𝑿 = 𝑿 = 𝑿 𝟏, 𝑿 𝟐, … , 𝑿 𝒏 𝑻 13
  • 14. 3.For the expressed criteria to be satisfied, certain condition about the noise- that affects our system- should be assumed: 𝑬 𝒘 𝒌 = 𝟎 𝑬 𝒗 𝒌 = 𝟎 𝑬 𝒘 𝒌 𝒗 𝒌 = 𝟎 𝒘 𝒌~(𝟎, 𝑸 𝒌) 𝑬 𝒘 𝒌 𝒘𝒋 𝑻 = 𝟎 𝒇𝒐𝒓 𝒌 ≠ 𝒋 𝑬 𝒗 𝒌 𝒗𝒋 𝑻 = 𝟎 𝒇𝒐𝒓 𝒌 ≠ 𝒋 𝒗 𝒌~(𝟎, 𝑹 𝒌) 4.Definition: 𝑸 𝒌= 𝑬 𝒘 𝒌 𝒘 𝒌 𝑻 : Covariance of w 𝑹 𝒌 = 𝑬(𝒗 𝒌 𝒗 𝒌 𝑻) : Covariance of v 14
  • 15. Standard Kalman filter algorithm: Step 0: Initialization of the filter at k=0. 𝒙 𝟎 + = 𝑬 𝒙 𝟎 𝑷 𝟎 + = 𝑬 𝒙 𝟎 − 𝒙 𝟎 + 𝒙 𝟎 − 𝒙 𝟎 + 𝑻 Step 1: for k=1,2,… follow the following procedure. a) 𝒙 𝒌 − = 𝑨 𝒌−𝟏 𝒙 𝒌−𝟏 + + 𝑩 𝒌−𝟏 𝒖 𝒌−𝟏 𝑷 𝒌 − = 𝑨 𝒌−𝟏 𝑷 𝒌−𝟏 + 𝑨 𝒌−𝟏 𝑻 + 𝑳 𝒌−𝟏 𝑸 𝒌−𝟏 𝑳 𝒌−𝟏 𝑻 b) 𝑲 𝒌 = 𝑷 𝒌 − 𝑯 𝒌 𝑻 𝑯 𝒌 𝑷 𝒌 − 𝑯 𝒌 𝑻 + 𝑴 𝒌 𝑹 𝒌 𝑴 𝒌 𝑻 −𝟏 c) 𝒙 𝒌 + = 𝒙 𝒌 − + 𝑲 𝒌 𝒚 𝒌 − 𝑯 𝒌 𝒙 𝒌 − 𝑷 𝒌 + = (𝑰 − 𝑲 𝒌 𝑯 𝒌)𝑷 𝒌 − 15
  • 16. Example on standard Kalman filter for linear systems (Vehicle Navigation): we want to model a vehicle going in a straight line. We can say that the state consists of the vehicle position p and velocity v. The input u is the commanded acceleration and the output y is the measured position. we are able to change the acceleration and measure the position every T seconds. In this case, elementary laws of physics say that the position p and velocity v will be governed by the following equations: 𝒑 𝒌 = 𝒑 𝒌−𝟏 + 𝑻𝒗 𝒌−𝟏 + 𝟏 𝟐 𝑻 𝟐 𝒖 𝒌−𝟏 + 𝒑 𝒌−𝟏 𝒗 𝒌 = 𝒗 𝒌−𝟏 + 𝑻𝒖 𝒌−𝟏 + 𝒗 𝒌 Where 𝑝 𝑘 and 𝑣 𝑘 are position and velocity noise respectively. Now we can define a state vector x that consists of position and velocity: 𝒙 𝒌 = 𝒑 𝒌 𝒗 𝒌 16
  • 17. Finally, knowing that the measured output is equal to the position, we can write our linear system equations as follows: 𝒙 𝒌 = 𝟏 𝑻 𝟎 𝟏 𝒙 𝒌−𝟏 + 𝑻 𝟐 𝟐 𝑻 𝒖 𝒌 + 𝒘 𝒌 𝒚 𝒌 = 𝟏 𝟎 𝒙 𝒌 + 𝒛 𝒌 𝑧 𝑘 is the measurement noise due to such things as instrumentation errors. If we want to control the vehicle with some sort of feedback system, we need an accurate estimate of the position p and the velocity v. In other words, we need a way to estimate the state x. This is where the Kalman filter comes in. 17
  • 18. The position is measured with an error of 10 feet (one standard deviation). The commanded acceleration is a constant 1 foot/sec2. The acceleration noise is 0.2 feet/sec2 (one standard deviation). The position is measured 10 times per second (T = 0.1). Since T = 0.1, the linear model that represents our system can be derived from the system model presented earlier as follows: 𝒙 𝒌 = 𝟏 𝟎. 𝟏 𝟎 𝟏 𝒙 𝒌−𝟏 + 𝟎. 𝟎𝟎𝟓 𝟎. 𝟏 𝒖 𝒌 + 𝒘 𝒌 𝒚 𝒌 = 𝟏 𝟎 𝒙 𝒌 + 𝒛 𝒌 Because the standard deviation of the measurement noise is 10 feet, the 𝑅 𝑘 matrix is simply equal to 100. Now we need to derive the 𝑄 𝑘 matrix. 18
  • 19. Since the position is proportional to 0.005 times the acceleration, and the acceleration noise is 0.2 feet/sec2, the variance of the position noise is (0.005)2× (0.2)2= 10−6. Similarly, since the velocity is proportional to 0.1 times the acceleration, the variance of the velocity noise is (0.1)2× (0.2)2= 4 × 10−4. Finally, the covariance of the position noise and velocity noise is equal to the standard deviation of the position noise times the standard deviation of the velocity noise, which can be calculated as 0.005 × 0.2 × 0.1 × 0.2 = 2 × 10−5. We can combine all of these calculations to obtain the following matrix for 𝑄 𝑘 : 𝑸 𝒌 = 𝑬 𝒙𝒙 𝑻 = 𝑬 𝒑 𝒗 𝒑 𝒗 = 𝑬 𝒑 𝟐 𝒑𝒗 𝒑𝒗 𝒗 𝟐 = 𝟏𝟎−𝟔 𝟐 × 𝟏𝟎−𝟓 𝟐 × 𝟏𝟎−𝟓 𝟒 × 𝟏𝟎−𝟒 I simulated the Kalman filter for this problem using Matlab. The results are shown in the accompanying figures. 19
  • 20. Figure 1 shows the true position of the vehicle, the measured position, and the estimated position. 20
  • 21. The two smooth curves are the true position and the estimated position, and they are almost too close to distinguish from one another. The noisy-looking curve is the measured position. For better comparison, part of the figure is magnified here: 21
  • 22. Figure 2 shows the error between the true position and the measured position, and the error between the true position and the Kalman filter‟s estimated position. 22
  • 23. Figure 3 shows the True and Estimated Velocity 23
  • 24. Figure 4 shows the error between the true velocity and the Kalman filter‟s estimated velocity. 24
  • 25. Shift to Nonlinear Systems: We have discussed state estimation for linear systems. But what if we want to estimate the states of a nonlinear system? As a matter of fact, almost all real engineering processes are nonlinear. Some can be approximated by linear systems but some cannot. This was recognized early in the history of Kalman filters and led to the development of the “extended Kalman filter” which is simply an extension of linear Kalman filter theory to nonlinear systems. A Nonlinear system is simply a process that can be described by the following two equations: 𝒙 𝒌 = 𝒇 𝒌−𝟏 𝒙 𝒌−𝟏, 𝒖 𝒌−𝟏 , 𝒘 𝒌−𝟏 𝒚 𝒌 = 𝒉 𝒌 𝒙 𝒌, 𝒖 𝒌 , 𝒗 𝒌 f : System function h : Output function 25
  • 26. Extended Kalman filter Algorithm: Step 0: Initialization of the filter at k=0 𝒙 𝟎 + = 𝑬 𝒙 𝟎 𝑷 𝟎 + = 𝑬 𝒙 𝟎 − 𝒙 𝟎 + 𝒙 𝟎 − 𝒙 𝟎 + 𝑻 Step 1: for k=1,2,… follow the following procedure. a) 𝑭 𝒌−𝟏 = 𝝏𝒇 𝒌−𝟏 𝝏𝒙 𝒙 𝒌−𝟏 + 𝑳 𝒌−𝟏 = 𝝏𝒇 𝒌−𝟏 𝝏𝒘 𝒙 𝒌−𝟏 + b) 𝒙 𝒌 − = 𝒇 𝒌−𝟏 𝒙 𝒌−𝟏 + , 𝒖 𝒌−𝟏, 𝟎 26
  • 27. 27 𝑷 𝒌 − = 𝑭 𝒌−𝟏 𝑷 𝒌−𝟏 + 𝑭 𝒌−𝟏 𝑻 + 𝑳 𝒌−𝟏 𝑸 𝒌−𝟏 𝑳 𝒌−𝟏 𝑻 c) 𝑯 𝒌 = 𝝏𝒉 𝒌 𝝏𝒙 𝒙 𝒌 − 𝑴 𝒌 = 𝝏𝒉 𝒌 𝝏𝒗 𝒙 𝒌 − d) 𝑲 𝒌 = 𝑷 𝒌 − 𝑯 𝒌 𝑻 𝑯 𝒌 𝑷 𝒌 − 𝑯 𝒌 𝑻 + 𝑴 𝒌 𝑹 𝒌 𝑴 𝒌 𝑻 −𝟏 e) 𝒙 𝒌 + = 𝒙 𝒌 − + 𝑲 𝒌 𝒚 𝒌 − 𝒉 𝒌 𝒙 𝒌 − , 𝟎 𝑷 𝒌 + = (𝑰 − 𝑲 𝒌 𝑯 𝒌)𝑷 𝒌 −
  • 28. 28 Example on Extended Kalman filter for Nonlinear systems: Consider a second order system as below: 𝒚 𝒖 = 𝝎 𝒏 𝟐 𝒔 𝟐 + 𝟐𝜻𝝎 𝒏 𝒔 + 𝝎 𝒏 𝟐 Define 𝑏 = 2𝜁𝜔 𝑛 We want to estimate 𝝎 𝒏 𝟐. In terms of differential equations: 𝒚 + 𝒃𝒚 + 𝝎 𝒏 𝟐 𝒚 = 𝝎 𝒏 𝟐 𝒖 And so we define the states as: 𝒙 𝟏 = 𝒚 𝒙 𝟏 = 𝒙 𝟐 𝒙 𝟐 = 𝒚 𝒙 𝟐 = −𝒃𝒙 𝟐 − 𝒙 𝟏 𝒙 𝟑 + 𝒙 𝟑 𝒖 𝒙 𝟑 = 𝝎 𝒏 𝟐 𝒙 𝟑 = 𝟎
  • 29. 29 Assume that: 𝝎 𝒏 = 𝟏 : true value 𝜻 = 𝟎. 𝟏 : zeta 𝑸 = 𝟏𝟎𝟎𝟎 𝟎 𝟎 𝟎. 𝟏 : Covariance of process noise 𝑹 = 𝟏𝟎 𝟎 𝟎 𝟏𝟎 : Covariance of measurement noise The simulation results are enclosed in the accompanying figures.
  • 30. 30 This figure shows the true and estimated value of 𝜔 𝑛 2
  • 31. 31 Variance of estimation error is implemented in this figure
  • 32. 32 References: 1. J. S. Meditch. Stochastic optimal linear estimation and control 2. Simon J. Julier and Jeffrey K. Uhlmann. A new extension of the Kalman filter to nonlinear systems. 3. Mohinder S. Grewal and Angus P. Andrews. Kalman filtering: Theory and Practice using Matlab 4. Rachel Kleinbauer . Kalman filtering Implementation with Matlab 5. Greg Welch and Gary Bishop. An introduction to the Kalman filter