Synthetic Power Quality Disturbances Dataset of Single and Combined Disturbances Generated in Accordance with IEEE 1159 Specifications

Citation Author(s):
Marwadi University
Marwadi University
Marwadi University
Marwadi University
Submitted by:
Last updated:
Wed, 03/06/2024 - 04:22
Data Format:
0 ratings - Please login to submit your rating.


The dataset encompasses a diverse array of electrical signals representing Power Quality Disturbances (PQD), both in single and combined forms, meticulously generated in adherence to the IEEE 1159 guideline. With a total of 15 different classes, each class contains 500 signals, summing up to a grand total of 7500 signals.

Each signal in the dataset is sampled at a frequency of 5 kHz, equivalent to 100 samples per cycle, ensuring precise capturing of waveform characteristics. The fundamental frequency across all signals is standardized at 50 Hz, reflecting typical power system frequencies. Additionally, all signals maintain a consistent amplitude of 1 per unit (pu), guaranteeing uniformity throughout the dataset.

The length of each signal is set to 1000 samples, equivalent to 10 cycles of the fundamental frequency, providing ample data for in-depth analysis while maintaining computational efficiency.

Crucially, the dataset includes both raw data and corresponding labels, facilitating supervised learning tasks and enabling the development and evaluation of classification algorithms. Researchers and practitioners in the field of electrical engineering can leverage this dataset to explore a myriad of topics, including signal processing, anomaly detection, fault diagnosis, and system resilience assessment, among others. The adherence to the IEEE 1159 guideline ensures that the dataset conforms to industry standards, enhancing its relevance and applicability in real-world scenarios.


PQData: This file comprises 7500 electrical signals, each containing 1000 samples. The signals represent various Power Quality Disturbances (PQD), including single and combined disturbances, generated in accordance with the IEEE 1159 guideline. Each signal is sampled at a frequency of 5 kHz, providing 100 samples per cycle. The fundamental frequency of all signals is set to 50 Hz, and each signal has a consistent amplitude of 1 per unit (pu). This dataset is suitable for tasks such as signal processing, anomaly detection, and fault diagnosis in electrical engineering applications.

PQLabel: This file contains labels corresponding to the 7500 signals in the signals dataset. The labels categorize each signal into one of the 15 predefined classes, allowing for supervised learning tasks such as classification. Researchers and practitioners can use this dataset in conjunction with the signals dataset for developing and evaluating classification algorithms, facilitating accurate identification and analysis of Power Quality Disturbances.