Byron Mason

DYNAMIC OPTIMISATION OF ELECTRIC VEHICLE POWERTRAINS Innovate UK and HORIBA (January 2024)

Application of mathematical optimisation to the calibration of an electric vehicle battery management system (BMS) and powertrain controller for improved energy efficiency. This project makes use of machine learning models to characterise the behaviour of systems within the vehicle. The models will be used within a multi-objective dynamic optimisation framework where the decision variables are the calibration parameters that influence battery and vehicle control.

  ONLINE ADAPTIVE OPTIMISATION AND MODEL PREDICTIVE CONTROL OF THERMAL ENERGY SYSTEMS EPSRC and Jaguar Land Rover (October 2023)

Development of thermal systems Model Predictive Controller (MPC) to reduce the overall energy consumption of the vehicle. The MPC takes account of the dynamics of the vehicle thermal systems and manages the distribution of thermal energy within the vehicle for improved performance. The novel control will be implemented using rapid control prototyping on dSPACE computational hardware.

  SYSTEM IDENTIFICATION USING BAYESIAN OPTIMISATION Caterpillar Inc (December 2021)

Physical modelling of systems within off-highway machines to help understand physical degradation processes and improve longevity of operation. Model parameters are identified within a Bayesian optimisation framework that makes the best use of existing knowledge whilst ensuring further testing maximises what is known about the system. This reduces wasted testing and ensures that best use of the testing time available is made. A significant part of this project was experimental and included the use of 400kW AC dynamometer, fast emissions measurement and a significant program of data acquisition.

  MACHINE LEARNING FOR BATTERY SYSTEM CHARACTERISATION Innovate UK and HORIBA (January 2021)

In this project machine learning models were used for characterisation of electric vehicle battery performance. The key to the success of the approach was excitation signal design to ensure that when data is gathered for model building it is fit for purpose. Novel excitation signal design was implemented experimentally to excite the battery dynamics leading to a more accurate understanding of battery performance.

  ELECTRIC MACHINE ADVANCED TESTING FACILITY EPSRC (January 2021)

Funding from the Engineering and Physical Sciences Research Council to install an advanced electric machine test facility. A particular challenge in this case was a peak power limitation on the local grid. This was overcome through design and the facility is currently in operation for various projects and it is also used in teaching.

  POWERTRAIN CHARACTERISATION AND OPTIMISATION Caterpillar Inc (September 2019)

Characterisation of power systems typically takes a significant amount of testing and is therefore time consuming and expensive. This project developed and implemented methods that resulted in a significant reduction in the amount of time taken to complete testing when compared with more conventional approaches. This gives the opportunity to take more data and improve what is known about the system and/or reduce product development time and cost. The project involved a large amount of experimental work using a 400kW AC dynamometer.

  DIGITAL ENGINEERING FOR AUTOMOTIVE PRODUCT DEVELOPMENT Innovate UK (January 2019)

Development of digital processes for vehicle engineering in partnership with several OEMs and suppliers. The main aim of the project was to use digital engineering processes to reduce costs and shorten development times. This covered multiple aspects from controls development to vehicle level testing and validation. Results included a significant reduction in hardware physical prototyping and through more thorough evaluation of various design options, discovery of issues earlier that can be resolved at significantly reduced cost.