olufemi olayiwola | Obafemi Awolowo University (original) (raw)
Uploads
Papers by olufemi olayiwola
A hybrid method comprising artificial neural networks (ANN) and discrete wavelet transform (DWT) ... more A hybrid method comprising artificial neural networks (ANN) and discrete wavelet transform (DWT) for short term load forecasting (STLF) is presented in this paper. For the ANN component of the hybrid model, the Levenberg-Marquardt back-propagation algorithm was used for network training; for the DWT component, the Daubechies mother wavelet was used. The hybrid ANN-DWT technique, the autoregressive integrated moving average (ARIMA) and the artificial neural network (ANN) techniques for STLF were developed in MATLAB 2016a and tested on real data obtained from the City of Cape Town Network Control Centre and the South African Weather Services. All three models were tested and compared for forecasting accuracy. The ranges of MAPE values for ARIMA, ANN, and the ANN-DWT hybrid model show the effectiveness of the hybrid ANN-DWT model in obtaining more accurate forecast results.
—The growth in photovoltaic installations makes characterization and condition monitoring essenti... more —The growth in photovoltaic installations makes characterization and condition monitoring essential. In this paper, broadband impedance spectroscopy is implemented for characterization and performance monitoring of silicon solar cells for near real-time operation. An optimized quasi-logarithmic broadband signal is designed to estimate the impedance response of the cells. Electrochemical equivalent circuits of the frequency response are then modelled from the obtained Nyquist plots and the cell parameters are extracted using complex nonlinear least squares. This procedure can be applied for direct estimation of the internal parameters of the silicon solar cells/module at different operating points. Results show that the implemented broadband characterization yields good correlation to the conventional electrochemical impedance spectroscopy at significantly reduced procedural time and equipment cost.
This paper presents a comparative analysis of the harmonics generated in the output waveforms of ... more This paper presents a comparative analysis of the harmonics generated in the output waveforms of 3-phase Hbridge cascaded multi-level inverter topology using Sinusoidal Pulse Width Modulation (SPWM) techniques. The Alternate Phased Opposite Disposition-Pulse Width Modulation (APOD-PWM) is one of the best SPWM schemes in achieving reduction in the Total Harmonic Distortion (THD), which is the major parameter often used to assess the suitability of an inverter. However, due to the closely set carrier arrangement, the APOD-PWM produces high THD especially at switching frequencies above 1-kHz. In this paper, a Gapped APOD-PWM is proposed. The switching sequence is similar to the APOD-PWM technique, but it has smaller switching angle. This reduces the switching overlap at boundary positions and hence the THD. The technique was simulated using the MATLAB-SIMULINK toolbox. The output current and voltage signals were evaluated and the THD was computed using SIMULINK-FFT tool. Results show that the proposed technique is able to further reduce THD at all switching frequencies.
A hybrid method comprising artificial neural networks (ANN) and discrete wavelet transform (DWT) ... more A hybrid method comprising artificial neural networks (ANN) and discrete wavelet transform (DWT) for short term load forecasting (STLF) is presented in this paper. For the ANN component of the hybrid model, the Levenberg-Marquardt back-propagation algorithm was used for network training; for the DWT component, the Daubechies mother wavelet was used. The hybrid ANN-DWT technique, the autoregressive integrated moving average (ARIMA) and the artificial neural network (ANN) techniques for STLF were developed in MATLAB 2016a and tested on real data obtained from the City of Cape Town Network Control Centre and the South African Weather Services. All three models were tested and compared for forecasting accuracy. The ranges of MAPE values for ARIMA, ANN, and the ANN-DWT hybrid model show the effectiveness of the hybrid ANN-DWT model in obtaining more accurate forecast results.
—The growth in photovoltaic installations makes characterization and condition monitoring essenti... more —The growth in photovoltaic installations makes characterization and condition monitoring essential. In this paper, broadband impedance spectroscopy is implemented for characterization and performance monitoring of silicon solar cells for near real-time operation. An optimized quasi-logarithmic broadband signal is designed to estimate the impedance response of the cells. Electrochemical equivalent circuits of the frequency response are then modelled from the obtained Nyquist plots and the cell parameters are extracted using complex nonlinear least squares. This procedure can be applied for direct estimation of the internal parameters of the silicon solar cells/module at different operating points. Results show that the implemented broadband characterization yields good correlation to the conventional electrochemical impedance spectroscopy at significantly reduced procedural time and equipment cost.
This paper presents a comparative analysis of the harmonics generated in the output waveforms of ... more This paper presents a comparative analysis of the harmonics generated in the output waveforms of 3-phase Hbridge cascaded multi-level inverter topology using Sinusoidal Pulse Width Modulation (SPWM) techniques. The Alternate Phased Opposite Disposition-Pulse Width Modulation (APOD-PWM) is one of the best SPWM schemes in achieving reduction in the Total Harmonic Distortion (THD), which is the major parameter often used to assess the suitability of an inverter. However, due to the closely set carrier arrangement, the APOD-PWM produces high THD especially at switching frequencies above 1-kHz. In this paper, a Gapped APOD-PWM is proposed. The switching sequence is similar to the APOD-PWM technique, but it has smaller switching angle. This reduces the switching overlap at boundary positions and hence the THD. The technique was simulated using the MATLAB-SIMULINK toolbox. The output current and voltage signals were evaluated and the THD was computed using SIMULINK-FFT tool. Results show that the proposed technique is able to further reduce THD at all switching frequencies.