Search

Unit commitment (UC) is a problem to economically determine the commitment status and generation level for each thermal unit during each time interval of the scheduled periods. The purpose of UC is to minimize the total generation cost while satisfying system demand, reserve requirements, and other unit constraints. The combined-cycled units have been increasingly installed throughout the world because of their high efficiency and fast response. However, the combined-cycle units usually have multiple operating configurations. The multi-configuration characteristic causes that the UC problem is more complex and difficult to be solved. In this paper, the authors present a simplified combined-cycled unit model to efficiently solve the related mixed integer linear programming-based UC problem. By simulating two test systems, results show that the proposed model is effective to reduce the complexity of problem with losing little solution accuracy.

Due to the increasing use of nonlinear loads, the associated power quality (PQ) problems have become a major concern to many electric utilities and their customers. For efficient and accurate analysis, the performance of computer simulation has been recognized as an important and necessary procedure to the power quality study. With the advanced development of computer and communication technologies, the real-time simulation becomes matured and provides better computational efficiency, flexible and scalable programming, and preserves a preciseness as good as conventional off-line simulation. This paper presents an application example of real-time simulation to harmonics study of an industrial power system. A cluster-based real-time simulator is executed under MATLAB/SIMULINK environment, where an industrial power system including the arc furnace and motor drives, as well as PQ mitigation devices is under test by real-time simulation techniques. In addition, the conventional off-line simulation with variable-step solver is included to compare the performance of the real-time simulation. Results indicate that the real-time simulation has a great advantage in both the computational efficiency and the system size under study than the off-line simulation, while the solution accuracy is maintained.

Fluorescent lamps are increasingly used in daily life, in which the gas discharge is the lighting principle. Due to the internal nonlinear characteristics of the lamp tube, harmonic currents are produced when the voltage source is directly connected to the lamp. Furthermore, in order to maintain the stable output of the lighting device, the ballast circuit is used. In practice, accurate fluorescent lamp models for harmonic studies are of importance to understand harmonic pollutions caused by fluorescent lamps. This paper presents a fluorescent lamp dynamic conductance model by using cubic spline interpolation method. To show the usefulness of the proposed model, a 10-W conventional 60Hz fluorescent lamp and a 20-W high-frequency fluorescent lamp are under tested. The laboratory experiments are also performed to verify the accuracy of the proposed model.

IEEE Std 519-1992 is regarded as one of the most widely used power quality standards for both electric utilities and their customers. Among several difficult aspects of applying IEEE Std 519, two issues are related to modeling and measurement of harmonic currents and voltages produced by nonlinear loads. This paper intends to discuss ambiguous aspects of the modeling and measurement when applying the standard.

Field measurements of voltage and current is the most effective way for characterizing the electric response of an EAF that describe the nonlinear behavior of AC EAF loads. Sufficient measured information can be adopted to develop an appropriate EAF model. In this paper, two classic methods based on measured data, harmonic current injections and equivalent harmonic voltage sources, for the EAF load modeling are reviewed. For comparison, two advanced methods based on actual recorded data, cubic spline interpolation and radial basis function neural network (RBFNN), are also proposed to model the EAF load. A steel plant power system with EAF loads is used for field measurements and computer simulations. Comparisons between the results of measured data and simulations for the four EAF models are being made according to the voltage/current waveforms and voltage-current characteristics. It is shown that the advanced models yield better performance than classic models of the EAF.