ASM Ashraf Mahmud

Dr. ASM Ashraf Mahmud

Computing, University of the West of Scotland, Paisley, Renfrewshire, UK
orcid
Lecturer (Assistant Professor)

Research Focus & Metrics

Academic Career Length

11Years (2016-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 April 2017
Verified
Conference Paper
Closed access
ASM Ashraf MahmudASM Ashraf Mahmud
ASM Ashraf Mahmud
PSPaul Sant
Paul Sant
The smart grid is the solution for the outdated current state of the power grid. An Energy Provider (EP) and Energy User (EU) are both looking to achieve savings. A Real-Time (RT) Demand Response (DR) pricing mechanism can reduce an EUs' bill and at the same time reduce the overall peak demand and Peak to Average Ratio (PAR) so that EP save their cost. In this paper, we will provide a model that uses a Price Suggestion Unit (pSU) and Price Control Unit (PCU) that provides Real-Time Price suggestions in the SG; we tested the model with a simultaneous perturbation stochastic approximation (SPSA) algorithm considering the threshold energy consumption with real data by using a stochastic iterative process. The EU has a choice to respond to the price suggestion. However, if they are non-responsive still they can still save on their bill in terms of taking advantage of the overall responses of other EUs.

14 June 2022
Verified
Conference Paper
Closed access
ASM Ashraf MahmudASM Ashraf Mahmud
ASM Ashraf Mahmud
The Smart Grid (SG) technologies greatly enhance the electrical grid by providing more energy security, increasing efficiency, reliability and economically benefitting the supplier which translates to reduce prices to the end customer. This is significant, as most of the current available energy comes from non-renewable sources. Economic development of nations result in increased demand for energy resources and as the world moves forward, the rate at which these resources are being spent is increasing. The SG poses numerous challenges in the fields of communication, engineering and policymaking. An aspect of the SG Technologies like Non-Intrusive Load Monitoring (NILM) enables the collection of valuable information that can be used to increase energy efficiency and gain deep insight on energy statistics. Novel software approaches like Deep Neural Networks, Markov models and Support Vector Machines play an important role for solving the NILM problem. In addition, researchers are delving into other aspects of the implementation such as privacy, price, and resource usage that coming from the paradigms of cloud and edge computing. Expert analysis on all of these topics is essential for the adoption of SG technology. In this paper, the merits of the various AI/ML techniques are expanded upon in a literature review. Then the previous and current state-of-the-art methods in the field of Deep Learning are evaluated and a conclusion is reached on their performance and viability. Furthermore, a case is made about relevant computing paradigms and how they may impact fundamental areas of the technology.

24 October 2016
Verified
Conference Paper
Closed access
ASM Ashraf MahmudASM Ashraf Mahmud
ASM Ashraf Mahmud
PSPaul Sant
Paul Sant
Faisal TariqFaisal Tariq
Faisal Tariq
DJDavid Jazani
David Jazani
The smart grid promises a myriad of benefits for both the consumer and energy service providers. However, realising its potential is subject to solving a number of complex issues. One of the major directions of smart grid research is demand response modelling that aimed at reducing the peak demand and billing by introducing appropriate real time pricing. The main difficulty is in managing the optimum pricing on a real time basis. In this paper, we will provide a state of the art review of an existing approach that models demand response in real time and also the underlying model that is tested with real data by using a stochastic iterative process with simultaneous perturbation stochastic approximation algorithm. Our model shows that real time price is better than a flat rate price. We will provide a brief outlook to the future where we propose a model that takes complex customers' behaviour into consideration. The proposed model includes a real time Price Suggestion Unit (PSU) that assists users to further reduce their electricity price while reducing the aggregate load for the energy providers.

7 August 2025 in Applied Sciences
Verified
Journal Article
Open access
PDF file
AEAFRIN JAHAN EVA
AFRIN JAHAN EVA
SASahal
Sahal
RARabita Amin
Rabita Amin
MKMuhammad R. A. Khandaker
Muhammad R. A. Khandaker
RKRisala Tasin Khan
Risala Tasin Khan
ASM Ashraf MahmudASM Ashraf Mahmud
ASM Ashraf Mahmud
Intelligent reflecting surface (IRS) is a promising technique which aims to shift the paradigm of uncontrollable wireless environment to a controllable one by adding the function of reconfigurability using multiple passive reflecting elements. In this work, optimal beamforming design for maximising achievable rate with respect to variable location of the IRS is considered. In particular, a single-cell wireless system that employs an IRS to aid communication between the user and an access point (AP) equipped with multiple antennas is adopted. An optimisation problem is formulated which aims to maximise the achievable rate, subject to signal-to-interference-plus-noise ratio (SINR) constraint of each individual user as well as the total transmit power constraint at the AP. The problem is solved by jointly optimising the transmit beamforming using active aerial array at the AP and the reflection coefficients using passive phase shifting at the IRS. Since the original optimisation problem is strictly non-convex, the problem is solved optimally by solving a corresponding power minimisation problem. Rigorous simulations have been carried out and the results demonstrate that the IRS-enabled system outperforms benchmark systems and employs significantly fewer RF power amplifiers.

3 August 2017 in EAI Endorsed Transactions on Energy Web
Verified
Journal Article
Open access
PDF file
ASM Ashraf MahmudASM Ashraf Mahmud
ASM Ashraf Mahmud
PSPaul Sant
Paul Sant
Faisal TariqFaisal Tariq
Faisal Tariq
DJDavid Jazani
David Jazani
The smart grid is the next generation bidirectional modern grid. Energy users’ are keen on reducing their bill and energy suppliers are also keen on reducing their industrial cost. Our demand response model would benefit them both. We have tested our model with the UK based traditional price value using a real-time basis. Energy users significantly reduced their bill and energy suppliers reduced their industrial cost due to load shifting. The Price Control Unit (PCU) and Price Suggestions Unit (PSU) utilise and embedded algorithms to vary price based upon demand. Our model makes suggestions based on energy threshold and makes use of stochastic approximation methods to produce prices. Our results shows that bill and peak load reductions benefit both the energy provider and users. This model also addresses users’ preferences, if users are non-responsive, they can still reduce their bills.

16 May 2026 in Scientific Reports
Verified
Journal Article
Open access
JSJulkar Nain Siam
Julkar Nain Siam
TSTanvir Ahsan Showrov
Tanvir Ahsan Showrov
SHSakir Hossain
Sakir Hossain
NANajmus Shakif Ayaan
Najmus Shakif Ayaan
SBS. M. Sadakatul Bari
S. M. Sadakatul Bari
ASM Ashraf MahmudASM Ashraf Mahmud
ASM Ashraf Mahmud
Motor imagery (MI)-based brain–computer interfaces (BCIs) enable users to control external devices using EEG signals, offering great potential in assistive and rehabilitation technologies. However, MI recognition remains challenging due to EEG’s low signal-to-noise ratio (SNR), inter-subject variability, and complex spatiotemporal patterns. Existing approaches often suffer from limited accuracy, high computational cost, and poor interpretability. In response to these challenges, we present the first comprehensive benchmarking of the publicly available EEG-hand movement (EEG-HM) dataset. Our study aims to establish a standardized performance baseline, guide the selection of optimal models by jointly considering accuracy, prediction time, and explainability, and ultimately accelerate progress in MI-BCI development. We have proposed a two-stage optimization of machine learning models that employs both feature selection and hyperparameter tuning. We exploit five feature selection algorithms for selecting the best set of EEG electrodes and frequency bands, while Bayesian optimization is exploited for machine learning model optimization through hyperparameter tuning. Furthermore, to validate the neurophysiological basis of our model’s decisions, we leverage explainable AI (XAI) algorithms—LIME and SHAP—quantifying the contributions of specific EEG electrodes and frequency bands to interpret its decision-making process. Through extensive simulations, the proposed two-stage optimization of the machine learning model demonstrates a superior performance in terms of accuracy, precision, and recall. This method outperforms the existing methods by 21.47% in accuracy with competitive prediction time. Its performance is further evaluated on the PhysioNet MI dataset, achieving a 4.67% accuracy improvement over state-of-the-art methods. Through LIME and SHAP, we provide the local and global explanations for no activity, left-hand, and right-hand imagery movements. Additionally, we analyze how various EEG frequency bands and electrode locations interact during the performance of different motor imagery hand movements.