Go to the Study here section
Go to the International section
Go to the Research section
The School aims to help individuals realise their full potential by developing their capacity to identify potential solutions across disciplines/ professions and overcome the challenges of rapidly changing environments. Our research degree programmes build on the research expertise within the Faculty of Technology, Design and Environment.
The primary aim of this investigation is the development of a machine learning based methodology that extracts unknown dynamics and physical laws from systems that exhibit complex/nonlinear and high dimensional traits. This will then be implemented in multi-objective experimental designs and control structures where observability (sensors) and controllability (actuators) are also limiting factors.
Current challenges in this area associated to a lack of interpretability, limited feature extraction, and the number of required data points are known to restrict the application of data driven machine learning to real engineering control and experimental design situations. Optimization techniques encompass a large portion of this problem, with system dynamic identification described above also being highly interlinked to methods for model order reduction, feature extraction, pattern recognition, defining control rules etc. Relevant cases within the experimental/research stages of the engineering process, such as design of experiment, test rig design and calibration, will be focused on.
Consideration will then be given to its intended implementation within a finished control system (controller rules, actuator/sensor impact/fidelity, dynamic Vs pre-recorded data, etc). This will manifest initially with a focus on system dynamics identification using deep learning and ensemble frameworks with simplified test cases. Ultimately a modern direct injection gasoline engine will be utilized as a validation case study for challenging dynamic behaviour. This will allow the exploration of limitations, simplifications and model-reductions that may be required to achieve implementable/realistic performance improvements and/or calibrations given practical factors such as analogue-digital transfer, reduced data points, system complexity, etc.
My PhD study is part of the Dynamic Modelling (DynaMo) Project in which Ford, Siemens, Oxford Brookes and other partners are united to study and minimise Internal Combustion Engine Soot emissions. My research project is focused on the CFD modelling of a modern small capacity direct injection turbocharged gasoline engine with all the processes that lead to the production of pollutants. After the injection of gasoline, air mixes with the fuel to create a homogenous mixture ready to burn. When this does not happen correctly, particulate matter forms as result of unwell premixed combustion. Studying the phenomena that take part to it is the key to in order to understand how to prevent soot particles to appear and be released in the atmosphere. In the Engine Modelling Team, my role involves also being responsible for the automatization of CFD model calibration and optimization.
Investigating techniques on data capture to allow monitoring of driver/vehicle performance and behaviour giving clear direction for design/system development. My general interests are mainly orientated around solving real time simulations that can be used for analysis or control purposes. I enjoy designing and developing accurate cost effective solutions that can help predict and explain real world behaviour through empirical techniques.
Electric motorport vehicles are often subjected to extremely harsh conditions during operation with drivers pushing the vehicles to the limit at every possible turn. Current trends in automotive electrification include improvements in vehicle range, performance, and reliability which all provide a cutting edge benefit to electric motorsport vehicles. My research is aimed towards improving vehicle range and performance through optimal real-time controls and high-performance Li-Ion cell modelling. These improvements lead to vehicle performance through improved strategies such as optimal energy deployment and dynamic power limitation.
I am also heavily involved in the High Voltage and Energy Storage Group, with our research aims to expand Oxford Brookes’ expertise in single cell and pack testing, pack design, and electrochemical modelling. Our lab’s focus is on acquiring cell data and using it to develop models in COMSOL, FeNiCs and MATLAB/Simulink. These models are then reduced, integrated with an optimal control architecture, and deployed on embedded hardware for hardware-in-the-loop (HIL) testing. Verification is done at the pack level with validation completed in vehicle to confirm on track performance.
The aim of my research is to understand the fragmentation mechanism of primary intermetallics formed in Al alloys using ultrasound-induced cavitation bubble dynamics. Research in solidification processing has recently been fast-tracked by the metal industries considering the benefits of producing lighter, stronger material in a sustainable, economical and pollution-free manner. The resulting production of high-quality light alloys are of great interest to the casting, automotive and aerospace industries. Cavitation melt treatment is one such eco-friendly and cost-effective alternative technology to conventional melt processes that can be applied to a range of typical and advanced metallic materials and is ready to be implemented in traditional casting technologies such as direct- chill (DC), continuous, or shape casting with the purpose of achieving high strength light alloys. This project ‘UltraMelt2’ is driven by the technological need to optimise and upscale cavitation melt processing technology. The project has been funded by Engineering and Physical Sciences Research Council (EPSRC), UK and is in collaboration with Brunel University, University of Greenwich, Anton Paar TriTec SA and Constellium.
This project particularly focuses on the random loading fatigue condition problems. Random vibration fatigue occurs in automotive, aerospace, offshore, wind turbine, structural and in machine components. The failure analyses of these types of issues are often dealt either using time domain or frequency domain fatigue analysis methods. The time domain approach used along with Miner’s rule of linear damage and rainflow counting algorithm is generally accepted as a comparatively good method random fatigue damage calculation. The frequency domain approach which uses the probability distribution functions delivers a faster alternative for the fatigue damage calculation, but the results are conservative and does not account for mean stress effect compared to the results obtained using time domain method Among the methods available for the frequency domain fatigue analysis, very few include the effect of mean stress but they are not widely accepted and some still need to be verified.
The main aim of the project is to develop a novel variable amplitude fatigue model using ANN, which include the effect of mean stress and give a better agreement with the time domain rainflow counting method. The developed model covers wide range of materials with ultimate tensile strength between 200 to 2000 MPa. The proposed model is generally applicable and faster ANN model for a wide range of random fatigue loading problems including the effect of mean stress.
Digital twins have become a crucial part of the design and development of internal combustion engines due to the reduced costs and increased speed to obtain optimal solutions that aim for efficiency gain and harmful emissions reduction. In this framework, the proposed project aim to use Computational Fluid Dynamic (CFD) analysis to develop robust semi-empirical models that correlate engine control variables to both air-fuel mixture parameters and soot emissions. This PhD project is attached to the APC DynAMO Collaborative Research and Development project (Innovate UK Grant N. 113130).
During the first phase of the PhD, a novel calibration methodology will be developed to quickly and accurately predict, within a commercial 3D CFD software, the fundamental characteristics of high- pressure fuel spray in a quiescent vessel. The subsequent implementation within the CFD model of the combustion chamber will reveal the characteristics of air-fuel mixture preparation and fuel to wall interactions.
The final phase of the project will focus on developing surrogate models of mixture preparation and Particle Matter (PM) emissions as well as optimal spray control strategies. New injection strategies will rely on injection timing, fuel mass delivered and injection pressure to improve the air-fuel mixture and minimise the fuel film formation. Since the injection phase influences considerably the combustion phase, the performance of the engine will be analysed for each new strategy to find a trade-off between PM emission cut-off and engine overall efficiency.
Formula Student is an exclusive competition to universities and colleges from around the world, to design and build a single seat race car to push the technical and practical ability of the students involved. Started in 2006, the V-Twin project aims to create the first student designed and built powertrain systems bespoke for the series, building on the success and heritage of the university’s team. The purpose of my research project is to complete the decade-long development of the engine and its advanced control systems: creating real data from which these systems can be iteratively adjusted through experimentation, this will improve the system’s performance and drivability for future vehicle installations. The broad range of research conducted, including control system design and calibration, extends published work and theories in the application of the V-Twin engine with additional resource constraints and complexities. The project helps to bridge the gap between theoretical understanding and practical testing, giving a wide range of opportunities and challenges to those involved. The formula student team has great links with industry and academia, providing a rare opportunity to make an impact on a project of such magnitude.
Drivers suffering from fatigue results in thousands of road accidents each year. The aim of my PhD is to develop artificial intelligence (AI) based techniques for fatigue detection using blink and facial expression rate as the main indicator. Most current technologies use image processing techniques to detect fatigue levels in real time. While effective it is inefficient when compared to Artificial Intelligence Techniques, specifically Neural Netowrks. While traditional image processing algorithms involve your program having to identify and locate facial features, Neural Networks are taught what an image is displaying using image labeling. Hence I will be achieving my research aim by developing a Neural Network that will be taught on how to identify if the driver is fatigued or distracted. There are three stages to a Neural Network. These are training, evaluation and testing. All stages are being carried out using a database of images faces of human faces. These images were taken while the participants of the database were driving.
The overall aim of the project is to develop a method of removing artefacts from biomedical signals that in the future could be used in bio-monitors, such as ECG machines, to aid interpretation and diagnosis by physicians. The project includes an investigation and analysis of the cause of artefacts (unwanted signals) and assess the current approaches in minimizing these artefacts. In addition, I will be designing active electrodes to maximize the signal to noise ratio of ECG signals. The proposed active electrodes will also be equipped with special electric circuits to capture body/electrode relative movement which is a major cause of artefacts. Moreover, I will simulate the design using software techniques and then, refine, build and test a working prototype of the electrodes. When the hardware is working and the signal to noise ratio is maximized, I will develop signal processing techniques to minimize artefacts. Finally, following artefact reduction, I will be presenting the data in a format (computer application) suitable for later display and diagnosis. I am well motivated by my supervisors and their continuous support and help. The research environment provided by the faculty is very helpful and besides that, the ability to access other libraries’ resources makes the research process smoother.