School of Aerospace, Transport and Manufacturing (SATM)

Permanent URI for this community

https://dspace.lib.cranfield.ac.uk/handle/1826/8736

Browse

Browsing School of Aerospace, Transport and Manufacturing (SATM) by Supervisor "Auger, Daniel J."

Now showing 1 - 4 of 4
  • Thumbnail Image
    ItemOpen Access
    Advanced state of charge estimation for lithium-sulfur batteries.
    (2017-05) Propp, Karsten; Auger, Daniel J.
    Lithium-sulfur (Li-S) batteries have a high theoretical energy density, which could outperform classic Li-ion technology in weight, manufacturing costs, safety and environmental impact. The aim of this study is to extend the research around Li-S through practical applications, specifically to develop a Li-S battery state of charge (SoC) estimation in the environment of electrical vehicles. This thesis is written in paper based form and is organised into three main areas. Part I introduces general topic of vehicle electrification, the framework of the research project REVB, mechanisms of Li-S cells and techniques for SoC estimation. The major scientific contribution is given in Part II within three studies in paper-based form. In Paper 1, a simple and fast running equivalent circuit network discharge model for Li-S cells over different temperature levels is presented. Paper 2 uses the model as an observer for Kalman filter (KF) based SoC estimation, employing and comparing the extended Kalman filter, the unscented Kalman filter and the Particle filter. Generally, a robust Li-S cell SoC estimator could be realized for realistic scenarios. To improve the robustness of the SoC estimation with different current densities, in Paper 3 a fast running online parameter identification method is applied, which could be used to improve the battery model as well as the SoC estimation precision. In Part III, the results are discussed and future directions are given to improve the SoC estimation accuracy for a wider range of applications and conditions. The final conclusion of this work is that a robust Li-S cell SoC estimation can be achieved with Kalman filter types of algorithms. Amongst the approaches of this study, the online parameter identification approach could deliver the best results and also contains most potential for further improvement.
  • Thumbnail Image
    ItemOpen Access
    An integrated battery unit regulation strategy
    (Cranfield University, 2023-03) Gong, You; Auger, Daniel J.; Fotouhi, Abbas
    In the research community, hybrid battery systems (HBSs) employing dual battery chemistries have been proposed as a solution to address the suboptimal overall performance exhibited by most state-of-the-art single-chemistry battery systems used in electric vehicle (EV) ap- plications. Currently, the predominant approach for regulating power distribution among different battery chemistries in HBSs is to configure DC/DC converters. However, the cost and weight associated with this configuration pose a significant barrier to its practical application. To overcome these limitations, this project presents a novel HBS design that utilizes a discrete-switched structure combined with intelligent low-frequency switching algorithms to replace DC/DC converters. The discrete-switched structure offers a simpler system architecture and lower power electronics costs while still maintaining the power allocation functionality of DC/DC converters. The switching algorithms developed, en- compassing heuristic and model-predictive control algorithms, enable the switching of cells within battery strings based on battery status and power demands, facilitating effec- tive power management. Through simulations and experiments, the HBS equipped with intelligent algorithms effectively regulates power distribution among different batteries and ensures a broadly balanced state of charge. Moreover, the novel HBS configura- tion employing nickel cobalt manganese oxide (NCM) and lithium-sulfur (Li-S) batteries has been thoroughly investigated, encompassing the hardware structure and control algo- rithms. This design enables both a long-range capability and high-power performance in EV applications. It should be noted that this work assumed the usage of homogeneous cells and effective cell cooling. Future research endeavors will focus on exploring cell-to-cell variations and the development of corresponding thermal management systems.
  • Thumbnail Image
    ItemOpen Access
    Driver distraction detection using experimental methods and machine learning algorithms.
    (Cranfield University, 2020-02) Zhang, Zhaozhong; Velenis, Efstathios; Fotouhi, Abbas; Auger, Daniel J.
    Driver distraction causes numerous road accidents, which is approximately equal to 25% of the total crashes according to the reports by the National Highway Traffic Safety Administration. Warnings can be helpful to mitigate the risks caused by driver distraction. Previous studies on driver distraction detection have not sufficiently found relevant input features to filter insignificant information, thus limiting the improvement of efficiency. Moreover, the disadvantages of driving simulators and public roads pose a challenge in collecting suitable data for feature identification and comparisons of performance among driver distraction detection algorithms. While the previous research focuses on improving prediction accuracy, shortening the prediction time is critical in giving timely warnings to drivers. This thesis aims at detecting driver distraction, which could provide faster and accurate warnings to drivers. The developed method is implemented by cutting the redundancy and irrelevant information fed to the algorithms and instead selecting suitable algorithms that achieve the balance between the prediction accuracy and prediction time. Moreover, a closed testing field supplies an environment for collecting more accurate information to identify the relevant features and to determine suitable algorithms. In this study, open-source data and experimental data are used. The results show that a balance between the prediction accuracy and the prediction time is achieved by feeding the relevant features and using suitable machine learning algorithms (e.g. Decision Tree). Compared with existing state-of-the-art methods, the prediction accuracy of the method proposed in this study has reached approximately the same level. More importantly, the efficiency has improved, including reduced prediction time and fewer input features. Consequently, less computer storage is used.
  • Thumbnail Image
    ItemOpen Access
    Optimisation methods for battery electric vehicle powertrain.
    (2017-07) Othaganont, Pongpun; Auger, Daniel J.; Whidborne, James F.
    The battery electric vehicle (BEV) is considered to be one of the solutions for reducing greenhouse gasses and an alternative means of transportation. However, some current limitations such as higher powertrain costs, limited driving range and negative perceptions of that range, have reduced BEVs’ popularity. This thesis aims to improve the tank-to-wheel energy consumption of the BEV by presenting possible powertrain architectures and developing new tools for powertrain analysis. The study has two main objectives; the first is to evaluate different possible powertrain topologies. The selected topologies include the single-motor single-axle, the double-motor double-axle, the in-wheel-motor single-axle and the in-wheel-motor double-axle. Models of these powertrains have been modified from the Quasi-Static toolbox, using vehicle parameters from the Nissan Leaf and subject to state assumptions. The multi-objective optimisation method has been applied to establish the costs/benefits of energy consumption, acceleration performance and powertrain cost. The results show that each topology presents its own benefits as the in-wheel types are good at energy efficiency and drivability, while the cost of the powertrain is the major drawback. The non-in-wheel-motor vehicle provides sufficient energy efficiency and driveability with lower powertrain cost. The second objective is to evaluate a possible alternative tool for BEV powertrain modelling and optimisation. The tool consists of four methodologies: sensitivity analysis, differential flatness, the Chebfun computational tool and the multi-disciplinary optimisation method. The study presents a possible alternative optimisation tool which may perhaps benefit the designer. This new tool may not be as convenient as the previous one; however, the new tool may give the designer greater understanding and insight into the BEV powertrain.