E-book details

Optimal input signal design in control systems identification

Optimal input signal design in control systems identification

Wiktor Jakowluk

Ebook

The topics covered in this book are primarily focused on optimal input design for system identification considering both the sensitivity and the prediction-error criteria. The problems of linear and non-linear dynamic systems identification of integer and fractional orders are considered. The book presents a variety of problems related to the design of optimal input signals including trajectory and end-point constraints, variable initial conditions, free final time tasks, and problems with cost functions endpoint. For the first time, an optimal input design formulation and a numerical scheme for fractional order system identification are presented. The application-oriented input design framework for open and closed-loop system identification is also widely presented. This framework allows for linear and finite-dimensional parametrization of the input signal spectrum design under frequency-by-frequency constraints. The author hopes that this book will be of interest to engineers and researchers in the field of automatic control and system identification.

Preface .................................................................................................................................. 7
List of symbols and acronyms ......................................................................................... 13
1. System identification for control ................................................................................ 17
1.1. A background to optimal input design .............................................................. 17
1.2. Model predictive control ...................................................................................... 19
1.3. Iterative feedback tuning ...................................................................................... 21
1.4. Adaptive parameter estimation ........................................................................... 22
2. Optimal inputs for (ODE) system identification ...................................................... 25
2.1. Single Weighted Cost Function Method ............................................................ 25
2.2. Input friendliness factor and system identification efficiency ........................ 29
2.3. Free final-time problem formulation ................................................................. 31
3. Optimal inputs for (FO) system identification ......................................................... 35
3.1. Fractional-order optimal control problem ........................................................ 35
3.2. Fractional-order operator approximation ......................................................... 38
3.3. Optimal input design problem reformulation .................................................. 42
4. Model-based input design in system identification ................................................. 45
4.1. Prediction error method ....................................................................................... 46
4.2. Input spectrum design problem .......................................................................... 47
4.2.1. Partial correlation parametrization ........................................................... 47
4.2.2. Finite-dimensional parametrization ......................................................... 48
4.3. Application constraints ........................................................................................ 48
5. Least-costly input design for closed-loop system identification ........................... 53
5.1. Model parameter estimation ................................................................................ 53
5.2. Uncertainty in the parameter estimates ............................................................. 55
5.3. Uncertainty of frequency function estimates .................................................... 56
5.4. Spectrum representation ...................................................................................... 57
5.5. Quality constraints in ellipsoidal regions ........................................................... 58
5.6. Optimization problem formulation .................................................................... 59

6. Continuous system identification in the time domain ............................................ 65
6.1. First-order linear system identification ............................................................. 65
6.2. Plant-friendly identification of a first-order system ........................................ 70
6.3. Plant-friendly identification of a second-order nonlinear system ................. 75
6.4. A single degree of freedom linear system identification .................................. 82
6.5. Free final time LTI system identification ........................................................... 90
6.6. Conclusions ........................................................................................................... 98
7. Fractional-order optimal input design problem .................................................... 101
7.1. Fractional-order inertial system optimal control ........................................... 101
7.2. Experimental results for different system orders ............................................ 104
7.3. Conclusions ......................................................................................................... 109
8. Application-oriented input design for open-loop focused on MPC .................. 111
8.1. Cascade tanks system overview ......................................................................... 111
8.2. Cascade tanks model identification .................................................................. 114
8.3. Robust multi-tank system control .................................................................... 121
8.4. Conclusions ......................................................................................................... 124
9. Application-oriented input design for closed-loop with fixed controller .......... 127
9.1. The nonlinear cascaded double tank system application .............................. 127
9.2. Optimal input spectrum design-upper tank case study ................................. 129
9.3. Optimal input spectrum design – interacting tanks case study .................... 132
9.4. Conclusions ......................................................................................................... 134
Summary and suggestions for future work ................................................................. 135
References ........................................................................................................................ 141
List of Tables ................................................................................................................... 151
List of Figures .................................................................................................................. 153
Summary .......................................................................................................................... 157
Streszczenie ..................................................................................................................... 159

  • Title: Optimal input signal design in control systems identification
  • Author: Wiktor Jakowluk
  • ISBN: 978-83-68077-18-6, 9788368077186
  • Date of issue: 2024-10-23
  • Format: Ebook
  • Item ID: e_4470
  • Publisher: Oficyna Wydawnicza Politechniki Białostockiej