Sajawal Gul Niazi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Sajawal Gul Niazi has a total of 65 co-authors. Here are the top co-authors by number of shared publications:
- Hong-Zhong Huang
University of Electronic Science and Technology of China
6 shared publications - Ghulam Moeen Uddin
Department of Mechanical Engineering, University of Engineering & Technology, Lahore 54890, Pakistan
3 shared publications - Syed Muhammad Arafat
Department of Mechanical Engineering, University of Engineering and Technology, Lahore H9H5+5C7, Pakistan
3 shared publications - Waqar Muhammad Ashraf
The Sargent Centre for Process Systems Engineering, Department of Chemical Engineering, University College London, Torrington Place, London WC1E 7JE, UK
3 shared publications - Ahmad Hassan Kamal
Huaneng Shandong Ruyi (Pakistan) Energy Pvt. Ltd. Sahiwal Coal Power Complex, Sahiwal Punjab 57000, Pakistan
2 shared publications - Ahsaan Jamil
Huaneng Shandong Ruyi (Pakistan) Energy Pvt. Ltd. Sahiwal Coal Power Complex, Sahiwal, Punjab 57000, Pakistan
2 shared publications
Latest Publications
- 8 October 2026
- Published by Wiley in Quality and Reliability Engineering International
A Robust Double‐Stage Feature Selection and TCN‐iT‐KAN Hybrid Framework for Remaining Useful Life Prediction of Rolling Bearings
The accurate prediction of the remaining useful life (RUL) of rolling bearings remains a challenging problem because degradation representations are often influenced by noise, feature redundancy, and the limited run‐to‐failure data. To overcome these challenges, this study develops a leakage‐aware prognostic framework that integrates a robust double‐stage feature selection strategy with a hybrid parallel predictor. The prediction model combines a parallel temporal convolution network, an inverse‐style Transformer, and a Kolmogorov–Arnold network head (TCN‐iT‐KAN). After identifying the first prediction time, a multi‐domain candidate feature bank is constructed from the time, envelope, frequency, wavelet packet, and discrete wavelet transform domains. Then, a double‐stage feature selection procedure is performed on the training set only to preserve informative features and avoid data leakage between data splits. The selected feature sequences are then used as input to a parallel prediction model, where the TCN captures local and multi‐scale temporal patterns, the inverse Transformer branch models global dependencies between features, and the KAN head performs nonlinear regression. Experimental results on the XJTU‐SY rolling bearing dataset, using a predefined data split, show that the proposed framework generally outperforms the benchmark models. The findings from ablation studies, branch sensitivity analysis, and post‐hoc feature analysis further show that these improvements are due to the combined effect of train‐only feature screening, parallel modeling of temporal feature interactions, and the KAN regression head.
- 8 October 2026
- Published by Wiley in Quality and Reliability Engineering International
A Robust Double‐Stage Feature Selection and TCN‐iT‐KAN Hybrid Framework for Remaining Useful Life Prediction of Rolling Bearings
The accurate prediction of the remaining useful life (RUL) of rolling bearings remains a challenging problem because degradation representations are often influenced by noise, feature redundancy, and the limited run‐to‐failure data. To overcome these challenges, this study develops a leakage‐aware prognostic framework that integrates a robust double‐stage feature selection strategy with a hybrid parallel predictor. The prediction model combines a parallel temporal convolution network, an inverse‐style Transformer, and a Kolmogorov–Arnold network head (TCN‐iT‐KAN). After identifying the first prediction time, a multi‐domain candidate feature bank is constructed from the time, envelope, frequency, wavelet packet, and discrete wavelet transform domains. Then, a double‐stage feature selection procedure is performed on the training set only to preserve informative features and avoid data leakage between data splits. The selected feature sequences are then used as input to a parallel prediction model, where the TCN captures local and multi‐scale temporal patterns, the inverse Transformer branch models global dependencies between features, and the KAN head performs nonlinear regression. Experimental results on the XJTU‐SY rolling bearing dataset, using a predefined data split, show that the proposed framework generally outperforms the benchmark models. The findings from ablation studies, branch sensitivity analysis, and post‐hoc feature analysis further show that these improvements are due to the combined effect of train‐only feature screening, parallel modeling of temporal feature interactions, and the KAN regression head.
- Open Access
- 13 October 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15 (20) , 10972
Unsupervised Convolutional Transformer Autoencoder for Robust Health Indicator Construction and RUL Prediction in Rotating Machinery
Prognostics for rotating machinery, particularly bearings, encounter significant challenges in constructing reliable health indicators (HIs) that accurately reflect degradation trajectories, thereby enabling precise remaining useful life (RUL) predictions. This article proposes a novel integrated approach for predicting the RUL of bearings without manual feature engineering. Specifically, a sequential autoencoder integrating a convolutional neural network (CNN) and vision Transformer (Vi-T) is employed to capture the local spatial patterns and global temporal correlations of time-domain vibration signals. The Wasserstein distance is introduced to quantify the divergence between healthy and degraded signal embeddings, resulting in a robust HI metric. Subsequently, the derived HI is fed into a CNN-bidirectional long short-term memory-regressor with Monte Carlo dropout to provide RUL predictions and Bayesian uncertainty estimates. Experimental results from the Xi’an Jiao-Tong University bearing dataset demonstrate that the proposed method surpasses conventional techniques in HI construction and RUL prediction accuracy, demonstrating its efficacy for complex industrial systems with minimal data preprocessing.
- 13 October 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15 (20) , 10972
Open AccessUnsupervised Convolutional Transformer Autoencoder for Robust Health Indicator Construction and RUL Prediction in Rotating Machinery
Prognostics for rotating machinery, particularly bearings, encounter significant challenges in constructing reliable health indicators (HIs) that accurately reflect degradation trajectories, thereby enabling precise remaining useful life (RUL) predictions. This article proposes a novel integrated approach for predicting the RUL of bearings without manual feature engineering. Specifically, a sequential autoencoder integrating a convolutional neural network (CNN) and vision Transformer (Vi-T) is employed to capture the local spatial patterns and global temporal correlations of time-domain vibration signals. The Wasserstein distance is introduced to quantify the divergence between healthy and degraded signal embeddings, resulting in a robust HI metric. Subsequently, the derived HI is fed into a CNN-bidirectional long short-term memory-regressor with Monte Carlo dropout to provide RUL predictions and Bayesian uncertainty estimates. Experimental results from the Xi’an Jiao-Tong University bearing dataset demonstrate that the proposed method surpasses conventional techniques in HI construction and RUL prediction accuracy, demonstrating its efficacy for complex industrial systems with minimal data preprocessing.
- 31 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-13
Domain-Specific Dual Network With Unsupervised Domain Adaptation for Transfer Fault Prognosis Across Machines Using Multiple Source Domains
Deep learning-based fault prognostic methods often require extensive training data, which can be challenging to obtain due to time, cost, or safety constraints. However, labeled data from related but distinct machines, such as from accelerated degradation experiments, can offer valuable insights. Most transfer learning methods focus on single-source domain adaptation (DA) within the same machine, overlooking variability within the source domain and missing the opportunity to leverage multi-source historical data effectively for application on different machines lacking run-to-failure data. To address these limitations, a domain-specific dual network (DSDN) is proposed for transfer fault prognosis across heterogeneous machines using multi-source domain data. The DSDN consists of two feature extraction modules: domain-specific degradation feature extraction (DSDFE) and domain-specific high-level feature extraction (DSHLFE). These modules incorporate a deep convolutional neural network (DCNN) enhanced with multiple convolutional block attention modules (CBAMs) and a multi-layer perceptron (MLP). The outputs from modules are concatenated to generate domaininvariant representations, effectively capturing both time-series and time-frequency spectrum features. By incorporating the DA technique, this approach mitigates discrepancies between the source and target domains, enabling accurate remaining useful life (RUL). The proposed method enhances the extraction of robust domain-invariant features for each source-target domain pair separately, improving prognostic accuracy by mitigating negative transfer. The effectiveness of the DSDN is validated through extensive case studies on single-source and multi-source bi-directional transfer fault prognosis tasks across machines, utilizing bearing datasets from the IEEE PHM Challenge 2012 and XJTU-SY. A comparative analysis further highlights its superiority over existing methods.
- 31 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-13
Domain-Specific Dual Network With Unsupervised Domain Adaptation for Transfer Fault Prognosis Across Machines Using Multiple Source Domains
Deep learning-based fault prognostic methods often require extensive training data, which can be challenging to obtain due to time, cost, or safety constraints. However, labeled data from related but distinct machines, such as from accelerated degradation experiments, can offer valuable insights. Most transfer learning methods focus on single-source domain adaptation (DA) within the same machine, overlooking variability within the source domain and missing the opportunity to leverage multi-source historical data effectively for application on different machines lacking run-to-failure data. To address these limitations, a domain-specific dual network (DSDN) is proposed for transfer fault prognosis across heterogeneous machines using multi-source domain data. The DSDN consists of two feature extraction modules: domain-specific degradation feature extraction (DSDFE) and domain-specific high-level feature extraction (DSHLFE). These modules incorporate a deep convolutional neural network (DCNN) enhanced with multiple convolutional block attention modules (CBAMs) and a multi-layer perceptron (MLP). The outputs from modules are concatenated to generate domaininvariant representations, effectively capturing both time-series and time-frequency spectrum features. By incorporating the DA technique, this approach mitigates discrepancies between the source and target domains, enabling accurate remaining useful life (RUL). The proposed method enhances the extraction of robust domain-invariant features for each source-target domain pair separately, improving prognostic accuracy by mitigating negative transfer. The effectiveness of the DSDN is validated through extensive case studies on single-source and multi-source bi-directional transfer fault prognosis tasks across machines, utilizing bearing datasets from the IEEE PHM Challenge 2012 and XJTU-SY. A comparative analysis further highlights its superiority over existing methods.
- 7 November 2023
- Published by Elsevier in Journal of Power Sources
- Vol. 590
A novel positional encoded attention-based Long short-term memory network for state of charge estimation of lithium-ion battery
- 7 November 2023
- Published by Elsevier in Journal of Power Sources
- Vol. 590
A novel positional encoded attention-based Long short-term memory network for state of charge estimation of lithium-ion battery
- Open Access
- 26 September 2023
- Published by MDPI AG in Applied Sciences
- Vol. 13 (19) , 10701
Multi-Directional Displacement Threshold Energy and Crystal Irradiation Damage Model
Subject to intense high-energy particle irradiation, various effects manifest within a material. Specifically, when high-energy particles collide with lattice atoms in a crystal material, a sequence of interactions is set in motion, initiating irradiation effects. Using GaAs solar cells as an example, this study investigates how potential barriers in different directions of atoms in sphalerite structures affect atomic displacement and establishes a probability model for lattice atomic displacement under proton irradiation. By combining Markov chains, changes in displacement threshold energy with different crystal orientations are described, and a damage model for cascading collision relationships that cause irradiation effects is established. Finally, the new model is compared to classical models, and differences in defects caused by proton impacts at several energies on GaAs crystals are simulated.
- 26 September 2023
- Published by MDPI AG in Applied Sciences
- Vol. 13 (19) , 10701
Open AccessMulti-Directional Displacement Threshold Energy and Crystal Irradiation Damage Model
Subject to intense high-energy particle irradiation, various effects manifest within a material. Specifically, when high-energy particles collide with lattice atoms in a crystal material, a sequence of interactions is set in motion, initiating irradiation effects. Using GaAs solar cells as an example, this study investigates how potential barriers in different directions of atoms in sphalerite structures affect atomic displacement and establishes a probability model for lattice atomic displacement under proton irradiation. By combining Markov chains, changes in displacement threshold energy with different crystal orientations are described, and a damage model for cascading collision relationships that cause irradiation effects is established. Finally, the new model is compared to classical models, and differences in defects caused by proton impacts at several energies on GaAs crystals are simulated.
- Open Access
- 1 January 2023
- Published by Research Publishing Services
- p. 428-436
Bearing Health and Safety Analysis to Improve the Reliability and Efficiency of Horizontal Axis Wind Turbine
- 1 January 2023
- Published by Research Publishing Services
- p. 428-436
Open AccessBearing Health and Safety Analysis to Improve the Reliability and Efficiency of Horizontal Axis Wind Turbine
- Open Access
- 1 January 2023
- Published by Research Publishing Services
- p. 1792-1799
Prediction of Remaining Useful Life of Bearings using a Parallel Neural Network
- 1 January 2023
- Published by Research Publishing Services
- p. 1792-1799
Open AccessPrediction of Remaining Useful Life of Bearings using a Parallel Neural Network
- Open Access
Optimization of a 660 MWe Supercritical Power Plant Performance—A Case of Industry 4.0 in the Data-Driven Operational Management. Part 2. Power Generation
Modern data analytics techniques and computationally inexpensive software tools are fueling the commercial applications of data-driven decision making and process optimization strategies for complex industrial operations. In this paper, modern and reliable process modeling techniques, i.e., multiple linear regression (MLR), artificial neural network (ANN), and least square support vector machine (LSSVM), are employed and comprehensively compared as reliable and robust process models for the generator power of a 660 MWe supercritical coal combustion power plant. Based on the external validation test conducted by the unseen operation data, LSSVM has outperformed the MLR and ANN models to predict the power plant’s generator power. Later, the LSSVM model is used for the failure mode recovery and a very successful operation control excellence tool. Moreover, by adjusting the thermo-electric operating parameters, the generator power on an average is increased by 1.74%, 1.80%, and 1.0 at 50% generation capacity, 75% generation capacity, and 100% generation capacity of the power plant, respectively. The process modeling based on process data and data-driven process optimization strategy building for improved process control is an actual realization of industry 4.0 in the industrial applications.
Open AccessOptimization of a 660 MWe Supercritical Power Plant Performance—A Case of Industry 4.0 in the Data-Driven Operational Management. Part 2. Power Generation
Modern data analytics techniques and computationally inexpensive software tools are fueling the commercial applications of data-driven decision making and process optimization strategies for complex industrial operations. In this paper, modern and reliable process modeling techniques, i.e., multiple linear regression (MLR), artificial neural network (ANN), and least square support vector machine (LSSVM), are employed and comprehensively compared as reliable and robust process models for the generator power of a 660 MWe supercritical coal combustion power plant. Based on the external validation test conducted by the unseen operation data, LSSVM has outperformed the MLR and ANN models to predict the power plant’s generator power. Later, the LSSVM model is used for the failure mode recovery and a very successful operation control excellence tool. Moreover, by adjusting the thermo-electric operating parameters, the generator power on an average is increased by 1.74%, 1.80%, and 1.0 at 50% generation capacity, 75% generation capacity, and 100% generation capacity of the power plant, respectively. The process modeling based on process data and data-driven process optimization strategy building for improved process control is an actual realization of industry 4.0 in the industrial applications.