H. J. Lu

Forecasting Flicker Severity by Grey Predictor

Rapid voltage fluctuations in an electric power network may produce significant levels of flickers, which have negative impacts on human eyes and power system components. This paper proposes a grey predictor model for the forecast of the flicker severity level associated with operating a large electric arc furnace load. Actual measured flicker index data are adopted to implement the predictor model. Test results based on the proposed model are compared with another neural-network-based method. It shows that a more accurate forecast is achieved by using the proposed grey predictor model.

Development of power quality analysis platform for INER Microgrid

The deployment of ac microgrids including distributed energy resources (DERs) can reduce the investment cost of building new fossil-fuel based power plants, facilitate the carbon-oxide reduction, provide reactive power compensation and frequency regulation, and enhance the system stability. However, some undesired effects such as imbalance, voltage fluctuation, and harmonics accompanied with the operations of DERs and nonlinear loads may affect the power quality (PQ) within the microgrid system. It is thus crucial to maintain the power quality within the microgrid to a satisfactory level. This paper presents the development of a graphical user interface-based analysis platform to monitor the power quality of the INER AC microgrid.

An Advanced EAF Model for Voltage Fluctuation Propagation Study

This paper presents an improved dynamic voltage-current model of an ac electric arc furnace (EAF) for multiple operation stages and for the study of propagations of voltage fluctuations in a power network associated with the EAF in operating stages. The EAF model is developed based on wavelet-transform and neural-network-based methods by training actual measured EAF current and voltage data. The proposed EAF model is then implemented in an actual 161-kV power system for simulations. The voltage fluctuations at other buses due to the EAF in operation are then assessed. Simulation and measured results show that the improved hybrid EAF model is accurate and is suitable for voltage fluctuation assessment when the actual measurements in the power network are limited or not available while the EAF is in operation or before the similar types of EAFs are to be installed in the system.

A study of passive harmonic filter planning for an AC microgrid

A microgrid usually consists of small-scale of thermal generation sources and other distributed energy resources (DER) such as photovoltaic, wind power, and fuel cells to serve its loads. The microgrid can increase DER penetrations and operation reliability through appropriate control schemes and energy management. When the inverter-based DERs supply electric energy to nonlinear loads in the microgrid, harmonic currents are produced and resonances in the grid may occur and causes negative impacts on some critical loads or other equipment, as well as extra power losses. To mitigate the undesired harmonic effects and improve the microgrid performance, this paper presents a study of planning single-tuned passive harmonic filters for the microgrid. Simulation results show that the designed filters can effectively mitigate the harmonic currents in the microgrid.

Neural network-based approach for determining optimal reference compensation current of shunt active power filter

With the increasing use of non-linear loads, the harmonic problems have become of great concerns. One of the solutions to harmonics is the application of shunt active power filters (APFs). The APF reference compensation strategy proposed in this paper adopts the backpropagation neural network (BPNN) to train the control algorithm and calculate the reference compensation current under steady state and varying load levels. Simulation results obtained by using MATLAB/Simulink show that the proposed BPNN-based APF control strategy can effectively mitigate harmonic currents generated by the nonlinear load.

A hybrid model for forecasting wind speed and wind power generation

Forecasting of wind speed and wind power generation is indispensable for the effective operation of a wind farm and the optimal management of revenue and risks. Hybrid forecasting of time series data is considered to be a potentially effective alternative compared with the conventional single forecasting modeling approaches such as autoregressive integrated moving average (ARIMA) and artificial neural network (ANN). Hybrid forecasting typically consists of a classic prediction model for the linear component of a time series and a nonlinear forecast model for the nonlinear component. This paper presents a hybrid approach combining ARIMA and radial basis function neural network for forecasting wind speed and wind power. Results obtained by a case study show that the proposed method is suitable for short-term forecasting applications.