Estimation and Decision Systems
Keywords |
Classification |
Keyword |
OFICIAL |
Automation and Control |
Instance: 2021/2022 - 1S
Cycles of Study/Courses
Acronym |
No. of Students |
Study Plan |
Curricular Years |
Credits UCN |
Credits ECTS |
Contact hours |
Total Time |
M.EEC |
16 |
Syllabus |
2 |
- |
6 |
39 |
|
Teaching language
Portuguese
Objectives
At the end of this course, students are expected to acquire a solid base of basic knowledge to understand
the estimation and identification problems and the methods that today constitute the "state of the art" in these areas, as well as decision making in an environment of uncertainty.
Learning outcomes and competences
At the end of this UC, students are expected to:
1. Understand Stochastic Processes in discrete time.
2. Know deterministic-stochastic transfer function models of discrete-time systems (ARX, ARMAX, Box Jenkins).
3. Know how to plan identification experiments and estimate models from experimental data.
4. Understand Markov Processes and Bayes Networks.
5. Understand Stochastic Dynamic Programming
6. Know how to build and use Markov decision processes
Working method
Presencial
Pre-requirements (prior knowledge) and co-requirements (common knowledge)
- Control Theory
- Algebra
- Probability and Statistics
Program
Stochastic Processes in discrete time.
Noise models.
Deterministic-stochastic transfer function models of discrete time systems (ARX, ARMAX, Box Jenkins). Prediction, Parameter Estimation and Identification of Discrete Time System Transfer Function Models (Least Squares, Instrumental Variables, and Prediction Error Optimization Methods).
Markov processes. Bayes nets. Inference. Stochastic Dynamic Programming and
Markov Decision Processes. Decision Theory
Mandatory literature
Ljung, Lennart;
System identification. ISBN: 0-13-881640-9
Puterman, M. L.; Puterman, M. L. (2014). Markov decision processes: discrete stochastic dynamic programming. , John Wiley & Sons., 2014
Complementary Bibliography
Verhaegen, Michel and Verdult, Vincent;
Filtering and System Identification - A least squares approach, Cambridge University Press, 2007. ISBN: ISBN-13 978-0-521-87512-7
Paulo Jorge de Azevedo Lopes dos Santos;
Identificação de sistemas dinâmicos
Lopes dos Santos, Paulo; Perdicoúlis, T-P A; Novara, Carlo; Ramos, Jose; Rivera, Daniel;
Linear Parameter Varying Systems - New Developments and Trends, World Scientific, 2012. ISBN: 13-978-981-4355-44-5
Kumar, P. R., & Varaiya, P. ; Stochastic systems: Estimation, identification, and adaptive control. ., SIAM, 2015
Graham C. Goodwin;
Adaptive filtering. ISBN: 0-13-004069-X
Teaching methods and learning activities
- Theoretical Lectures: Subject exposition using slides and the board.
- TP Lectures: Execution of Lab works with real or simulated data to demonstrate of concepts. Small Identification, decision and Estimation Projects.
- Lectures will be of theoretical presentations illustrated with real and simulated examples using MATLAB/Octave
that clarify the concepts and results presented. Resolution of exercises and implemention of small projects,
proposed by the lecturers, that encourage an active and critical participation of students. Use of
computational tools in data processing namely MATLAB/Octave
Software
Octave
Matlab
System Identification Toolbox
keywords
Technological sciences > Engineering > Systems engineering > Systems theory
Evaluation Type
Distributed evaluation without final exam
Assessment Components
Designation |
Weight (%) |
Teste |
90,00 |
Participação presencial |
10,00 |
Total: |
100,00 |
Amount of time allocated to each course unit
Designation |
Time (hours) |
Estudo autónomo |
78,00 |
Frequência das aulas |
56,00 |
Trabalho laboratorial |
28,00 |
Total: |
162,00 |
Eligibility for exams
Do not exceed the absence limit.
Calculation formula of final grade
The students have to do 2 tests. The first one at the middle of the semester and the second one at the end.
They will also have to carry out regular questionnaires.
The final grade is calculated by the formula
final grade=0.5*T1+0.5*T2++0.1*Q
T1 -First Test
T2 - Second teste
Q - questionnaires
An oral exame is required for those students who want more than 18
Students with less than 10 may take the appeal examination. In this case the final grade will be the one obtained in the examination.
Special assessment (TE, DA, ...)
The students admitted to the examination because they were released from attending the lectures (according to points a) b) c) of Article 4 of the General Evaluation Rules), will make the ordinary witten tests.
Classification improvement
The classification improvement can be made in special examination period . The examination grade will be the final grade if better than the previous one.
Observations
Students who obtained frequency in the previous year (only in the previous year) .