APAS – Automated Powerunit Architecture Selection and Control Strategy Generation
This IDE funded project aims to develop capability to model and simulate powerunit systems and controllers for a variety of architectures. These models are then used to assess attribute performance, which in turn then optimises the vehicle architecture to meet a set of targets and requirements.
Collaborative Partners
McLaren Automotive (Lead)
University of Bath
Ricardo UK Ltd
​
IDE people
Monitoring Officer
£768k
total project spend
Two PhD students from the University of Bath will work in close connection with McLaren Automotive and Ricardo to:
-
Develop software that allows multiple attributes to be assessed for a single vehicle model, and the architecture to be modified based on these results.
​​
-
Develop enabling technologies and advanced control and optimisation approaches to allow for the optimum definition of a vehicle powertrain architecture.
Project Summary
The APAS project unlocked the opportunity to perform optimisation of a powertrain architecture based on multiple requirements like lap time, CO2 and acceleration performance.
The project focussed on a single architecture to develop the optimisation routines and developed methods to assess the effectiveness of different optimisation approaches.
​
A method was developed to automatically calibrate the torque split hybrid strategy for a given architecture. This was used in the optimisation routine to give a true like-for-like comparison.
​
The APAS project has:
​
-
Established a methodology for characterising powertrain components in a way that is useful to an optimiser
​
-
Defined a general optimisation process for powertrain component selection using Genetic Algorithms
​
-
Developed a tool for comparing Equivalent Consumption Minimization Strategy with Dynamic Programming methods for fuel economy optimisation
​
-
Delivered a software that McLaren can integrate in their pipeline to evaluate different alternatives in the future
An IDE Funded Project