| Summary: | The accuracy of the parameter identification of power system model and efficiency of frequency control are part of the challenging work in power system operation and control area. Whereas, the complexity and high non-linearty of the power system model have led to the continuing research for improvement that still extensively active, especially for load frequency control (LFC). Generally, LFC is responsible to maintain the zero steady-state errors in the frequency changing and restoring the natural frequency to its normal position. Many methods have been proposed and implemented in identification of power system and LFC, however, they may not be appropriate. For example, the classical methods for parameter identification (LSE and MLE), the classical methods for LFC (PI, PD and PID) and the intelligent methods (fuzzy logic, neural network, genetic algorithm, and PSO). Thus, motivated from the topics, this Thesis is brought to present the improvement of the parameter identification of power system model and the response of the LFC in power system. The Thesis is divided into two parts in accordance to the topic. Where, in the first part, the coherent identification algorithm for single and multi-area power systems with disturbances is proposed. A new method from the improvement of Particle Swarm Optimization (PSO) is developed in order to find the best global optimal value. Meanwhile, part two presents three developed control methods for FLC from the improvement of fuzzy control (named as scaled fuzzy using PSO, parallel conventional PI/PD with Scaled Fuzzy PI/PD and Mirror Fuzzy controller) by adapting the utilization of PSO to optimize the scaled gain of fuzzy controllers. These proposed control methods in LFC will be examined and verified in two and four areas power system. The outcomes of the proposed parameters identification and LFC control methods are presented the results through simulation using Matlab by making a comparison on the frequency transient response. Various analyses are shown and the discussions on the results are done appropriately. Lastly, the Thesis is given the concluding remarks and the contributions which can be specified into two, a modification of PSO for parameters identification named as PSO segmentation and a new fuzzy control named as a Mirror Fuzzy controller for LFC
|
|---|