Use this URL to cite or link to this record in EThOS:
Title: Perturbation estimation based control of electric energy conversion systems
Author: Ren, Yaxing
ISNI:       0000 0004 6422 5444
Awarding Body: University of Liverpool
Current Institution: University of Liverpool
Date of Award: 2017
Availability of Full Text:
Access from EThOS:
Access from Institution:
The traditional power system includes the centralised power generation, high voltage AC power transmission and three phase energy consumption. Electric energy conversion systems (ECSs) have been applied to the power generation, energy storage and power consumption to convert energy between the electric form and other forms. In the future power system, the ratio of distributed power generation and storage will have a rapid increment with the development of power electronics technology. Thus, the robustness and stability are significant to the ECSs in the future power electronics enabled power system. This thesis deals with the design and analysis with theoretical contribution, and the implementation of a perturbation estimation based nonlinear adaptive control (NAC) on ECSs, i.e. the wind turbine (WT), the energy storage system (ESS) in converter based microgrid (MG), and the induction motor (IM), respectively, in simulation and experimental validation. The wind turbine is one of the most promising distributed power generation resources. The challenge in controlling a wind turbine is its nonlinear behaviour of aerodynamics under random wind speed. This makes it difficult to obtain the optimal control performance operating under the time-varying wind speed via conventional linear control method. In addition, as the future power system including plenty of distributed generation and consumption, typically in MG application, the ESS is necessary to balance the power difference between power generation and consumption. Due to the low stiffness and inertia of an islanded MG, the challenge is the stability problem and power quality of MG under unknown disturbance and unbalanced power demand. Moreover, other than the disturbance from power generation side, plenty of unknown disturbance also appears on the power consumption. The most popular workhorse for industrial application is the induction motor (IM), which is affected by the disturbance of unknown load torque under operation. The IM has highly coupled states and nonlinear interactions between states. The conventional vector control depending on the flux position is sensitive to parameter changes. And the use of a speed encoder increases the risk in the IM speed drive in the electric vehicle application. To cope with these challenges in the ECS applications, the perturbation estimation based control method is studied and applied to improve the robustness of the ECSs for power generation, storage and consumption of the future power system. In the control method, a state and perturbation observer is used to estimate the perturbation term, which includes the nonlinear interactions between states, external disturbance, parameter changes and unmodeled dynamics. In the WT pitch angle control, a nonlinear PI-based controller is designed with a perturbation observer to estimate and compensate the system nonlinearities and disturbance of WT system. In the ESS voltage control of islanded MG, a voltage controller is designed for the ESS in MG via estimating and compensating the unknown disturbance to reduce the voltage unbalance rate. In the IM speed drive, an NAC based speed controller is investigated to control the IM directly under the stationary frame to improve its robustness under disturbance and parameter uncertainty. Another contribution is to propose a speed sensorless NAC controller with a combined SPO to control the IM without the dependency of a speed sensor. The proposed control methods are compared with the conventional methods regarding their control performance. The results show that the perturbation estimation based method can improve the robustness of ECS under disturbance and parameter uncertainty in the renewable power generation, MG bus voltage regulation, and IM speed drive. However, the great observer bandwidth can amplify the sensor noise and reduce the robustness and stability of the closed loop system. In the study, the observer and controller bandwidth is set greater than the controller bandwidth and lower than the sensor noise bandwidth, with optimised bandwidth tuned via pole placement method and the closed loop stability of the ECS systems is analysed using Lyapunov theory.
Supervisor: Jiang, Lin Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral