The Library
Control-oriented dynamical modelling and state estimation of centrifugal fans with induction motors drives
Tools
Turkeri, Cebrail (2023) Control-oriented dynamical modelling and state estimation of centrifugal fans with induction motors drives. PhD thesis, University of Warwick.
|
PDF
WRAP_Theses_Turkeri_2023.pdf - Submitted Version - Requires a PDF viewer. Download (4Mb) | Preview |
Official URL: http://webcat.warwick.ac.uk/record=b3986165
Abstract
To achieve significant energy savings in air and water supply systems, variable frequency induction motor drives are employed to power centrifugal fans and pumps. The cost of implementing closed loop control for energy savings can be reduced by replacing expensive flow rate or pressure sensors with estimators that are based on monitored variables from the drives and are embedded into their software. Existing estimation techniques rely on quasi-steady modelling of fans and pumps using data from steady state experiments or data sheets. However, this approach faces challenges during transients, which are addressed in the present thesis. This research proposes quasi-steady and dynamical estimators for the flow rate and pressure of a centrifugal fan with an induction motor drive, based on neural networks trained using both steady state and transient experimental data. The thesis describes a test rig designed specifically for this purpose, which utilizes an industrial centrifugal fan from Nicotra- Gebhardt equipped with a three-phase induction motor. Additionally, the thesis presents a detailed procedure for designing the estimators.
The use of electrical drives for controlling centrifugal fans and pumps is a wellestablished practice that can lead to significant energy savings. However, this requires electrical and automation engineers to possess knowledge relevant to the modelling of fans and pumps in relation to drives. Existing approaches rely heavily on quasisteady modelling, which is widely used by drives application experts, but there is limited adoption of dynamical modelling that is integrated with AC drives.
This thesis aims to enhance existing dynamical models by employing neural network estimation to calculate the overall efficiency of the fan/pump. The model separates the fan/pump's own efficiency for the computation of motor load torque to reflect accurately the power balance. The developed model is designed to be suitable for control design applications. Additionally, this research verifies the dynamical model experimentally, which supports the necessity of incorporating a first-order nonlinear differential equation to model the flow rate dynamics.
A linearized mathematical model of a centrifugal fan powered by a Squirrel Cage Induction Motor has been developed and validated through experimental and simulation data. Furthermore, a H mixed sensitivity approach has been employed to design a controller for a system that aims to achieve performance objectives such as stability, disturbance rejection, and reference tracking. The proposed H mixed sensitivity controller meets the necessary requirements for optimal performance, offering a promising solution for control system design.
Item Type: | Thesis (PhD) | ||||
---|---|---|---|---|---|
Subjects: | Q Science > QA Mathematics > QA76 Electronic computers. Computer science. Computer software T Technology > TK Electrical engineering. Electronics Nuclear engineering T Technology > TL Motor vehicles. Aeronautics. Astronautics |
||||
Library of Congress Subject Headings (LCSH): | Electric motors, Induction, Air flow -- Mathematical models -- Data processing, Fans (Machinery), Centrifugal pumps -- Blades, Sustainable engineering, Neural networks (Computer science) | ||||
Official Date: | April 2023 | ||||
Dates: |
|
||||
Institution: | University of Warwick | ||||
Theses Department: | School of Engineering | ||||
Thesis Type: | PhD | ||||
Publication Status: | Unpublished | ||||
Supervisor(s)/Advisor: | Kiselychnyk, Oleh | ||||
Sponsors: | Turkey. Millî Eğitim Bakanlığı | ||||
Format of File: | |||||
Extent: | xv, 154 pages : illustrations | ||||
Language: | eng |
Request changes or add full text files to a record
Repository staff actions (login required)
View Item |