Load frequency controller design based on extreme learning machine

The loading of a power system is never constant. The actual load change of the power system cannot be predicted at any point in time. A load change in any are of the power system will result in a change in frequency of the power system. Frequency is a major stability criterion for large-scale multi...

وصف كامل

محفوظ في:
التفاصيل البيبلوغرافية
المؤلف الرئيسي: Wu, Si.
مؤلفون آخرون: Wang Youyi
التنسيق: Theses and Dissertations
اللغة:English
منشور في: 2009
الموضوعات:
الوصول للمادة أونلاين:http://hdl.handle.net/10356/18882
الوسوم: إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
المؤسسة: Nanyang Technological University
اللغة: English
الوصف
الملخص:The loading of a power system is never constant. The actual load change of the power system cannot be predicted at any point in time. A load change in any are of the power system will result in a change in frequency of the power system. Frequency is a major stability criterion for large-scale multi area systems. To improve the stability of the power networks, it is necessary to design a load frequency control system. The designed controller must be able to cope with parametric uncertainties and nonlinearity of a real power system. In this dissertation, a load frequency controller based on the Riccati-equation approach designed by the author’s supervisor Dr. Wang Youyi will be introduced. Only the bounds of the system parameters are required to design this controller. Simulation results show that the robust load frequency controller can ensure that the system is stable for all admissible uncertainties, even in the presence of generation rate constraint. In the following part of the dissertation, it is proposed to use a neural network controller based on ELM (Extreme Learning Machine) algorithm instead of the robust load frequency controller. The performance data of the robust controller is set as the training pairs for the ELM neural-net load frequency controller. It aims to get a more adaptive control system for a larger parameters range. The performance of the ELM neural network load frequency controller is then compare with the original robust load frequency controller in Chapter 5 of this dissertation.