چكيده به لاتين
Electric power quality (PQ) is one of the most critical factors effective on the reliability and energy efficiency of power systems. With the development of microprocessors and power electronics in industrial installations, the electrical equipment sensitivity to PQ has dramatically increased. On the other hand, some power quality disturbances (PQDs) can lead to power outages or interruption in service, and equipment damage, which is very expensive and time-consuming. According to studies, most complaints about low power quality are related to short-term voltage and harmonic disturbances. Therefore, rapid and accurate detection of these disturbances is essential. In this dissertation, two different models have been used to detect short-term voltage power quality disturbances and harmonic distortion. In the first model, the variational mode extraction (VME) method based on data measured by micro-phasor measurement units (µPMUs) is used for detecting power quality disturbances. In this model, first, the optimal μPMUs placement is performed using the proposed improved hybrid differential evolutionary and genetic algorithm (DE-GA). Two scenarios with and without considering the effect of zero injection buses (ZIBs) in the optimal μPMUs placement (OPP) are considered to achieve the full network observability and the assurance of accurate measurement of the PQD by at least one device. Also, the bus observability Index (BOI) and system observability redundancy index (SORI) have been used to increase the network observability redundancy, in which case, if one µPMU is lost, the visibility of the buses is provided by another µPMU. In the second stage, the new variational mode extraction (VME) method is used to detect the single and combined PQ disturbances. The simulation results show that the VME method while separating the disturbances from the sinusoidal signals, is able to determine the type of disturbance, also it has a higher speed and less computation than other methods presented in different papers. Also, the accuracy of the VME method results is significantly increased in the single extraction of a particular mode of the signal. The advantage of using the proposed µPMU-based PQD detection method is obtaining phasor information and observing online network changes. Indeed, the superiority of the proposed method is due to the use of data and real-time measurements of µPMUs in the VME algorithm, which provides the ability to transfer information with very high accuracy and speed. In other words, the proposed method is an online real-time PQ detection system that effectively detects PQ disturbances in real-time applications.
In the second scenario of the proposed PQDs detection, a new estimation model is used to simultaneously extract the fundamental and harmonic phasors information of the system. Also, the recursive variational mode extraction (RVME) approach efficiency has been investigated for detecting power quality disturbances (PQDs). In the first step of this method, the optimal μPMUs placement (OPP) is done for complete observability of the studied systems. Then, the observability status of the system is checked. If a number of system buses are not observable for any reason (the single μPMU loss condition, measuring device failure, etc.), a proposed phasor estimation model based on measuring μPMUs is used to obtain other required data. In other words, the signals measured by μPMUs as well as the results of the proposed phasor estimation model will be used to obtain the data needed for PQDs detection. The proposed phasor estimation model is used to determine and separate fundamental and harmonic components information of voltage and current signals. Finally, the recursive variational mode extraction (RVME) approach has been applied for detecting short-term voltage disturbances and harmonic distortion in distribution networks. The RVME approach is an extended model of the VME method and has robust mathematical principles that can extract the modes correctly from each other. In other words, the RVME approach is used to extract and decompose an input signal into a discrete number of sub-signals (modes). Therefore, the PQDs detection is done by extracting the required voltage (current) signals by helping μPMUs data and the proposed phasor estimation model and using the recursive variational mode extraction method.
In this dissertation, to evaluate the effectiveness of the two proposed detection models mentioned above, six different scenarios of the single and combined PQ disturbances are investigated. These disturbances include four single disturbances (sag, swell, interruption, harmonic distortion) and two combined disturbances (sag with harmonics and swell with harmonics). To evaluate the performance of the proposed method in noisy conditions, Gaussian noise with a signal-to-noise ratio (SNR) of 30 dB has been used. In other words, to increase the robustness of the proposed method, Gaussian white noise was randomly added to the artificial PQD data at different levels. The obtained results confirm that the proposed method is resistant to noise and has high accuracy in detecting single and combined PQDs studied even in the presence of noise. The results obtained from various simulations, as well as using real data, indicate that VME and RVME approaches can extract a special mode of the signal and appropriately detect single and combined disturbances from other PQDs. Also, these methods have higher speed and fewer computations than other methods presented in different papers. On the other, the use of measuring μPMUs and the proposed phasor estimation model in the proposed PQDs detection methods have high accuracy and speed, which can provide the ability to observe dynamic changes in the network.
