Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.779825
Title: System identification and adaptive current balancing ON/OFF control of DC-DC switch mode power converter
Author: Wang, Chen
Awarding Body: Newcastle University
Current Institution: University of Newcastle upon Tyne
Date of Award: 2018
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
Reliability becomes more and more important in industrial application of Switch Mode Power Converters (SMPCs). A poorly performing power supply in a power system can influence its operation and potentially compromise the entire system performance in terms of efficiency. To maintain a high reliability, high performance SMPC effective control is necessary for regulating the output of the SMPC system. However, an uncertainty is a key factor in SMPC operation. For example, parameter variations can be caused by environmental effects such as temperature, pressure and humidity. Usually, fixed controllers cannot respond optimally and generate an effective signal to compensate the output error caused by time varying parameter changes. Therefore, the stability is potentially compromised in this case. To resolve this problem, increasing interest has been shown in employing online system identification techniques to estimate the parameter values in real time. Moreover, the control scheme applied after system identification is often called "adaptive control" due to the control signal selfadapting to the parameter variation by receiving the information from the system identification process. In system identification, the Recursive Least Square (RLS) algorithm has been widely used because it is well understood and easy to implement. However, despite the popularity of RLS, the high computational cost and slow convergence speed are the main restrictions for use in SMPC applications. For this reason, this research presents an alternative algorithm to RLS; Fast Affline Projection (FAP). Detailed mathematical analysis proves the superior computational efficiency of this algorithm. Moreover, simulation and experiment result verify this unique adaptive algorithm has improved performance in terms of computational cost and convergence speed compared with the conventional RLS methods. Finally, a novel adaptive control scheme is designed for optimal control of a DC-DC buck converter during transient periods. By applying the proposed adaptive algorithm, the control signal can be successfully employed to change the ON/OFF state of the power transistor in the DC-DC buck converter to improve the dynamic behaviour. Simulation and experiment result show the proposed adaptive control scheme significantly improves the transient response of the buck converter, particularly during an abrupt load change condition.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.779825  DOI: Not available
Share: