We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Dual Identification of Multi-Complex and Non-Stationary Power Quality Disturbances Using Variational Mode Decomposition in Hybrid Modern Power Systems.
- Authors
Behzadi, Mania; Askari, Mohammad Tolou; Amirahmadi, Meysam; Babaeinik, Majid
- Abstract
Power quality disturbances (PQDs) in hybrid modern power (HMP) systems are a challenging issue, particularly with the complex and non-stationary behaviour of disturbances generated by the power electronic components intrinsic in these technologies. The identification of PQDs and their sources is the first step to improve the power quality in HMP systems. The novelty of this paper comes from the accurate, fast, and dual identification of the multi-complex and non-stationary PQDs and their sources in the presence of different distributed generations and loads. This paper identifies the PQDs and their sources simultaneously using variational mode decomposition (VMD) and K-nearest neighbours classifier. The features are extracted using VMD, just from voltage waveforms. To reduce the redundant data, dimension of features vector, and time, the Relief-F method and correlation feature selection method are applied to the extracted features. To verify the effectiveness of the proposed method, different scenarios such as misfiring, variation of sun radiation and wind speed, entrance and exit of loads, capacitors and distributed generators, different faults at the grid in half-load to full-load were simulated. Furthermore, to verify the proposed method, our method was compared with PNN and the different distance functions of KNN. Hence, two feature selectors, Relief-F and CFS, were compared together. Results show that the accuracy and speed of the Cityblock of KNN are better than other distance functions and PNN. Moreover, the performance of Relief-F is better than CFS in this problem. Thus, to verify the proposed method in noisy conditions, three signal-to-noise ratios of 20 dB, 30 dB, and 50 dB are investigated. The obtained results confirm that the proposed method is robust to noise and has high accuracy in identifying the multi-complex and non-stationary PQDs and their sources even in noisy conditions. This method can be used as an added algorithm for smart metering in hybrid modern power systems in real time.
- Subjects
HYBRID power systems; FEATURE extraction; HYBRID power; POWER quality disturbances; DISTRIBUTED power generation; SOLAR radiation; DISTANCES; DOMAIN decomposition methods; FEATURE selection
- Publication
Arabian Journal for Science & Engineering (Springer Science & Business Media B.V. ), 2022, Vol 47, Issue 11, p14389
- ISSN
2193-567X
- Publication type
Article
- DOI
10.1007/s13369-022-06787-5