Concepts of identification of dynamic systems
Continuous and discrete time dynamic systems analysis. Block diagrams and feedback. Frequency response. Sampling theorem. Fourier transform.
Matrices, eigenvalues and eigenvectors, vector spaces.
Derivatives, integrals, gradients, limits, stationary points of a mathematical function.
Random variables, probability distributions, moments of a probability distribution.
The Adaptive Learning, Estimation and Supervision of dynamical systems (ALES) course is the natural continuation of the Model Identification and Data Analysis (MIDA) course, with the further topic of fault diagnosis, nowadays a fundamental part to implement predictive maintenance concepts from an industry 4.0 perspective.
At the end of the course, the student will be able to:
• Employ and implement recursive and adaptive estimation methods, without and with constraints on the estimation variables
• Estimate models of dynamical systems in state-space form and with multiple-inputs and multiple-outputs
• Estimate models of dynamical systems under closed-loop conditions
• Define and solve industrial fault diagnosis problems, identifying its main components and envisage a possible solution using a model-based approach
PART 1: RECURSIVE AND ADAPTIVE ESTIMATION
Recursive least squares method (RLS). Oblivion factor. Receding Horizon Estimation (RHE).
PART 2: SUBSPACE IDENTIFICATION
Estimation of a dynamic model in state space. N4SID and MOESP methods, with extension to MIMO systems.
PART 3: CLOSED LOOP IDENTIFICATION
PEM identification for closed loop systems. Comparison with open-loop identification.
PART 4: SUPERVISION AND DIAGNOSIS
Model-based fault diagnosis. Additive and multiplicative faults. Modeling of a dynamical system with faults. Parity space approach. Observer approach. RLS-based approach.
The educational path proposed to the student is as follows:
1) Follow the lecture presented with the slides. They are all available before the start of the course and it is useful for the student to view the slides of a lesson before following the lesson.
2) Study the topics of the lesson with the help of the textbook, slides and personal notes.
3) Implement the algorithms seen in class, independently, in MATLAB. The assessment of the implementation exercises provides points for passing the exam.
4) Towards the end of the lessons the student is able to carry out an exam example, proposed with its solution.
Much importance is given to students' active participation in lessons, which is stimulated through continuous dialogue. Students can find the teacher at any time (preferably by appointment) by going to the teacher's office (Office 303 Building C).
The exam consists of two parts:
1) 1 hour written exam with 3 open-ended theory questions. Each question is worth 5 to 8 points. The written exam is worth a maximum of 15 points. At least 10 points are required to pass the written exam and access the project evaluation.
2) Evaluation of a project with oral discussion. The project consists in implementing an algorithm from the scientific literature in MATLAB. The project is worth up to 15 points.
The exercises at home, carried out through MS Teams, contribute up to a maximum of +3 points on the final grade.
At the address https://cal.unibg.it you can find all the information on the course.
The materials for following the lessons and for personal preparation are made available on the Teams Group and on the course website.
In case of public authority actions for the containment of epidemiological emergencies, the teaching modality could undergo changes compared to what is stated in the syllabus, to make the course and exams in line with the sanitary limitations.