TY - THES A1 - Tahoon, Mohammad T1 - Estimating the State of Charge of Lithium-ion Batteries using Deep Learning for Electric Vehicle Applications N2 - Reliable and safe operation of Li-ion batteries in electric vehicles relies on an accurate estimation of their states, more specifically their state of charge (SOC), which is used for managing charging and discharging, for example. Deep learning is increasingly used for modeling complex systems, which generates more interest in battery state estimation. The theoretical investigation of candidate neural networks for SOC estimation is preceded an experimental approach to evaluate the shortlisted neural networks performance. Publicly available testing datasets representing real driving cycles has been used to train the selected networks, taking into consideration different operating temperatures for the exact cycles. Testing the prediction capability of the models across a realistic range of operating conditions is the driving factor for including testing cycles with a temperature range of [0–40] °C in the training datasets. A demonstration of the full training cycle is conducted, along with the optimization problem of network parameters. This work demonstrates the applicable neural network family used for training this kind of sequential data, followed by the shortlisting of two neural networks based on their performance using literature research. After selecting the networks, finer evaluation criteria is applied, including both their prediction capability and training requirements. The results show that both networks, long short-term memory (LSTM) and gate recurrent units (GRU), offer satisfactory prediction capability with mean absolute error (MAE) of (2.9%, 2.7%) respectively, when averaged across the tested temperatures. The GRU network on the other hand had significantly higher resource requirements and a more complex architecture. Finally, a more recent architecture is briefly discussed, borrowing the strength of a parallel application field. The scope of this work is to highlight both the capabilities and cost of training deep learning networks for SOC estimation. KW - Electric vehicles KW - Lithium-Ion batteries KW - State of charge KW - Deep learning KW - Neural networks Y1 - 2023 ER - TY - THES A1 - Cheung, Ho Yu T1 - Comparison of wheel and rubber track drives for tractors regarding masses, mass inertias, energy demands and tractor mechanics N2 - Soil compaction is a major concern in agriculture. As farming machinery develops to be larger and heavier, the risk of soil compaction increases. Tractors with rubber tracks have been developed to increase the contact area with soil in order to reduce the risk of soil compaction as well as to increase tractive performance. However, tractors with rubber tracks have increased mass due to the track components. Therefore, the combined effect of rubber tracks was investigated. Comparison of tyres and rubber tracks for tractors on mass, mass inertia, tractor mechanics, energy demand, and soil compaction were investigated. Configurations of tractors consisting of combinations of tyres and interchangeable rubber tracks are evaluated in calculations and in simulation tool Terranimo. Simplified rubber track models were built in order to investigate effect of inertia from the rotational components on energy demand. A theoretical scenario was constructed to calculate the energy demand for the predefined workload. Experimental data in past research showed that the superior tractive performance of rubber track is more apparent in looser soil. In the energy demand calculation part of the thesis the tracked configurations resulted in a higher energy demand. The tracked configurations also showed effect on reducing risk of soil compaction in the same defined scenario. However, the increased energy demand, hence higher fuel consumption and energy efficiency is a downside of the tracked configurations, and the selection between tyre and rubber track should depend on the situation. KW - Rubber tracks KW - Soil compaction KW - Tractor KW - Energy demand KW - Agriculture Y1 - 2022 ER - TY - THES A1 - Hammond, Luciano T1 - Feasibility check, design and dimensioning of a scrounger magnet coupling to a Siemens-E-Highway-truck N2 - Among other factors, electric vehicles (EVs) were developed with the intention of mitigating the transportation sector’s impact on the environment and the depletion of fossil fuels. While various methods exist to charge electric vehicles (EVs) or extend their driving ranges, each have their benefits and drawbacks. As such, a method to accomplish the latter, i.e., to extend the driving range of an EV – and more specifically, a battery electric vehicle (BEV) – is proposed, which entails magnetically coupling to an electric truck (ET) ahead. This is enabled by a front-mounted system, comprised of a magnetic circuit (MC) and a linear axis (LA), which acts as a drawbridge to extend the MC ahead of the vehicle to a steel plate located at the rear of the ET. By utilising an array of permanent magnets (PMs) and a solenoid within a pot ferrite core (PFC) structure, the proposed system is able to easily generate a coupling force of around 6400 𝑁, which is sufficient to maintain the magnetic couple to the ET ahead, even under slight deceleration of the BEV of 2 𝑚 ⁄ s^2, as it recuperates energy via regenerative braking. With the proposed system, a power gain of 30.24 𝑘𝑊 is achievable in this manner, which allows for a significant improvement to the driving range of BEVs, as well as providing similar benefits as those offered by dynamic charging systems, namely, the convenience, and is thus regarded as a feasible solution, with regard to its performance. KW - Master's thesis KW - Magnet coupling Y1 - 2022 ER - TY - THES A1 - Ganev, Ivan T1 - Design and Implementation of a Virtual Gearbox N2 - This thesis represents the design and implementation of a virtual gearbox functionality as a system with an end goal to be used on the Electric Gokart platform. The design includes a 3-part solution consisting of: - Human-machine interface (HMI) – that is a physical hardware as a separate unit with its own housing - Intermediary software for transfer of input information and output feedback of the user – as software code - Control Model – MATLAB Simulink model to be run on the main computer of the platform to facilitate the motion control Due to the system complexity and multidomain solution the work has been broken down and documented to focus on a specific aspect for each deliverable. The human machine interface focuses on the hardware aspect of the functionality – research on and evaluation of different components for the input, feedback and communication and the impact of each component to the further development of the intermediary software. The software part is focused on choosing and implementing an efficient and logical handling of the sampling and relay of information to prevent the potential problem of communication network overload. And finally, the Simulink Control Model will focus on investigating input information, processing and motion control realised via torque control. The information is to be used used in developing a system response model evaluating it on the prototype and allowing tuning via system identification in the MATLAB Simulink development environment. Y1 - 2025 ER -