## SYSTEM IDENTIFICATION AND FILTERING (8 ECTS) : 2017-2018

### SYLLABUS

- Basics on probability theory.

- Basics on estimation theory. Optimality criteria: centering, consistency, efficiency. Rao-Cramer lower-bound for covariance.

- Optimal estimation: least square estimates (WLS), maximal likelihood estimates (ML)  and Bayesian estimates (B).

- Kalman filtering and prediction (KF and KP). Steady state Kalman filter (SSKF) and asympotitc optimality.

- Parametric identification.  Identification with test inputs/control inputs.  Identification for input-output and state space models. Mixed ML-KF techniques of identification

- Autoregressive models: AR, MA, ARMA and ARMAX models.

### EXAMS

EXAM STRUCTURE: The exam consists of an oral part on the topics covered in the course (see the syllabus)  and the  discussion of a project  (see available projects below). Oral part and discussion of the project are done at the same time.

PROJECT DESCRIPTION: Each project is of a THEMATIC or PRACTICAL type.  THEMATIC projects concern specific theoretical topics studied and pioneered in scientific papers. PRACTICAL projects are solutions of identification  and filtering problems arising in practical applications.

PROJECT REPORTS: For thematic projects, a detailed written report must be produced on the assigned paper, proving a complete understanding of the technical solutions and simulations given in the paper and using detailed technical discussions and motivations. For practical projects, a detailed written report must be produced on the solution of the problem, proving a complete understanding of the mathematical tools adopted for the solution and one or more simulations to prove the effectiveness of the solution proposed.

PROJECT ASSIGNMENT: Each project is assigned to a maximum number of 2 students (two different reports  by each student or a unique report  by two students together). Once a project is chosen by two students (and approved for assignment) it is not available any more for assignment. A project assignment request is applied for by sending a mail to battilotti@diag.uniroma1.it including the name of the student, student number  and the title of the project. A notification for the assignment will be sent asap.

### AVAILABLE PROJECTS

THEMATIC PROJECTS

1) Extended Kalman filtering and weighted least squares dynamic  identification of robot (paper)*****NO MORE  AVAILABLE FOR ASSIGNMENT (assigned to #1817923, #1795683)

2) Conditional Particle Filters for Simultaneous Mobile Robot Localization and People-Tracking (paper)*****NO MORE  AVAILABLE FOR ASSIGNMENT  (assigned to #1596046, #1482987)

3) System identification application using Hammerstein model (paper)*****NO MORE  AVAILABLE FOR ASSIGNMENT  (assigned to #1673687, #1724744)

4) Methods for the Nonlinear Transformation of Means and Covariances in Filters and Estimators (paper)*****NO MORE  AVAILABLE FOR ASSIGNMENT (assigned to #1474502, #1769020)

5) Localization of Mobile Robot using Particle Filter (paper)*****NO MORE  AVAILABLE FOR ASSIGNMENT (assigned to #1794619, #1795330)

6) Unscented Kalman Filter (paper)*****AVAILABLE FOR ASSIGNMENT

7) Recursive Least Squares Estimation (paper) *****NO MORE  AVAILABLE FOR ASSIGNMENT (assigned to #1796868, #1805201)

8) Subspace identification (paper*****NO MORE  AVAILABLE FOR ASSIGNMENT (assigned to #1771834, #1804827)

9) Weighted and Generalized Least Squares (paper)*****NO MORE  AVAILABLE FOR ASSIGNMENT (assigned to #17793479, #1761125)

10) Overview of Maximum likelihood estimators (paper)*****NO MORE  AVAILABLE FOR ASSIGNMENT (assigned to #1313336, #1618107)

11) System Identification in Computer Vision (paper) *****NO MORE  AVAILABLE FOR ASSIGNMENT (assigned to #1652946, #1457275)

12) Nonlinear Identification Methods for Modeling Biomedical Systems (paper) *****NO MORE  AVAILABLE FOR ASSIGNMENT (assigned to #1821455, #1771702)

13) Stochastic Stability of the Discrete-Time Extended Kalman Filter (paper) *****AVAILABLE FOR ONLY ONE ASSIGNMENT (assigned to #ADRIJA SEN)

14) Autoregressive Moving Avarage models I (paper) *****AVAILABLE FOR ONLY ONE ASSIGNMENT (assigned to #1686128)

15) Autoregressive Moving Avarage models II (paper)*****AVAILABLE FOR ASSIGNMENT

16) Particle Filters for Online Nonlinear/Non-Gaussian Bayesian Tracking (paper)*****NO MORE  AVAILABLE FOR ASSIGNMENT  (assigned to #1797978, #1828140)

17) Monte Carlo Particle Filter (paper) *****NO MORE  AVAILABLE FOR ASSIGNMENT (assigned to #1772676, #1772167)

18) Particle Filters (paper) *****NO MORE  AVAILABLE FOR ASSIGNMENT (assigned to #1580884, #1823229)

19) Prediction Error Estimation Methods (paper)*****NO MORE  AVAILABLE FOR ASSIGNMENT (assigned to #1797529, #1771689)

20) Unscented Kalman Filtering for  spacecraft attitude state and parameter estimation (paper)*****NO MORE  AVAILABLE FOR ASSIGNMENT (assigned to #1827213, #1826995)

21) Particle filter for UAV trajectory prediction under uncertainties (paper)*****NO MORE  AVAILABLE FOR ASSIGNMENT (assigned to #1805845, #1794130)

PRACTICAL PROJECTS

1) KF and EKF for image processing *****NO MORE  AVAILABLE FOR ASSIGNMENT (assigned to #DHULIPALLA)

### STUDENT HOURS

Friday from 3 to 6 p.m. (Via Ariosto 25,  room A207)

***************************************************

SUPPLEMENTARY MATERIAL: Available Written Exams for past sessions (pdf files)