Multivariable sliding-mode extremum seeking control in power electronic systems

Date created: 
2019-08-20
Identifier: 
etd20529
Keywords: 
Multivariable extremum seeking
Sliding-mode
Maximum power point tracking
Alternator-based energy conversion
PI tuning
Permanent magnet synchronous motor
Multi-objective extremum seeking
Exercise machine
Abstract: 

This thesis investigates the design and implementation of extremum seeking control with application to power electronics. To this end, a novel multivariable sliding-mode extremum seeking (MSES) scheme is developed and applied to several control and optimization problems involving maximum power point tracking (MPPT) and motor drives. The behavior of the controller in terms of convergence characteristics and stability is studied using nonlinear systems analysis tools. The proposed MSES is utilized in three applications. First, we apply the concept to MPPT in an alternator-based energy conversion system. The objective is to achieve optimal power conversion at different speeds and output voltages of a Lundell alternator. The performance of the proposed controller is experimentally verified on a laboratory-scale setup through controlling the alternator field current and output voltage to gain fast and precise convergence and robust performance in face of disturbances and uncertainties. In the second application, the proposed MSES is used to tune a proportional-integral (PI) controller which regulates the current of a permanent magnet synchronous motor (PMSM). The performance of the proposed MSES tuning method in terms of accuracy, parametric variations, and load torque disturbances is investigated through several experimental tests on a PMSM setup. In the third application, the MSES concept is extended to a PMSM-drive system which emulates an exercise machine working at low speeds. In this case, the algorithm is modified to a multi-objective sliding-mode extremum seeking (MOES) optimization scheme for torque control of a PMSM as well as minimization of its torque ripples. To this end, the MSESC method is utilized to implement an adaptive iterative learning control (AILC) strategy for torque ripple minimization. The performance of the proposed MOES in terms of torque ripple suppression, steady state and transient performance, and load disturbance rejection is experimentally verified through synthesizing different mechanical impedances.

Document type: 
Thesis
Rights: 
This thesis may be printed or downloaded for non-commercial research and scholarly purposes. Copyright remains with the author.
File(s): 
Supervisor(s): 
Mehrdad Moallem
Department: 
Applied Sciences: School of Mechatronic Systems Engineering
Thesis type: 
(Thesis) Ph.D.
Statistics: