Revolutionary Electric Vehicle Battery (REVB) - design and integration of novel state estimation/control algorithms & system optimisation techniques

点击翻译按钮获取中文摘要

The latest draft of the main application form (submitted by OXIS Energy on behalf of all collaborators) is attached. Thisgives a fuller picture of the project, including detailed work package descriptions.Cranfield's research is divided into six work packages:WP CF1: High-Level Requirements GatheringWP CF2: Architectural DesignWP CF3: Modelling and EstimationWP CF4: Control and OptimizationWP CF5: IntegrationWP CF6: Development of Reusable Software ToolsDetailed descriptions of each follow.WP CF1: High-Level Requirements GatheringIn this work package, we will work with the other consortium members to determine the exact requirements that theautomotive battery pack needs to meet. We will understand the power requirements and other demands on the batterysystem.The ultimate end use of the technology would be in an electric vehicle. For the purpose of this project, the consortium isplanning to produce a hardware-in-the-loop 'technology demonstrator'. This will be developed by Lotus Engineering, andCranfield will use it to integrate and test state estimators and control algorithms. (The requirements for the hardware-in-the-loop simulator are directly analogous to those for a real electric vehicle, so we can be confident that our work has realworldrelevance.)WP CF 2: Architectural DesignIn this work package, we will work with one collaborator in particular (Lotus Engineering) to design the structure of thedemonstrator's Battery Energy Manager. (This is a computer control system that can be embedded in a vehicle, and wewill be using it to control the behaviour of the powertrain components.) We will first seek to understand Lotus's existingcontroller in detail. After this, we will be able to produce detailed requirements for our controller and estimator, and thendesign an architecture for it.WP CF3: Modelling and EstimationIn this work package, we will develop low-order models of the battery suitable for embedding in the Battery EnergyManager, and we will use these to design a novel state estimator that will give 'virtual measurements' for quantities that arehard to measure directly. We will prepare an initial version of these in time to support Lotus's parallel software designactivities. We will then refine the algorithms, taking into account development in our collaborators' research activities.The low-order models will also be used to design controllers in WP CP4.WP CF4: Control and OptimizationIn this work package, we will apply multi-objective system optimization techniques to the driveline as a whole. We will alsouse advanced control techniques to develop a novel control algorithm for the Battery Energy Manager. We will prepareinitial versions of these in time to support Lotus's parallel software design activities. We will then refine the algorithms,taking into account development in our collaborators' research activities.WP CF5: IntegrationIn this work package, we will support Lotus as they integrate our estimation and control algorithms into their Battery EnergyManager and hardware-in-the-loop technology demonstrator. We will simulate the behaviour in a virtual environment,modify the algorithms if needed, and then support integration on the hardware-in-the-loop technology demonstrator itself.WP CF6: Development of Reusable Software ToolsIn this work package, we will take the software tools we develop in earlier work packages, and develop them to make themrobust enough to be useful to others who wish to apply similar techniques. (We will distribute our tools over the WWW.)