Design optimisation and real-time energy management control of the electrified off-highway vehicle with artificial intelligence

Zhou, Quan ORCID: 0000-0003-4216-3468 (2019). Design optimisation and real-time energy management control of the electrified off-highway vehicle with artificial intelligence. University of Birmingham. Ph.D.

Preview

Zhou2019PhD.pdf
Text - Accepted Version
Available under License All rights reserved.
Download (7MB) | Preview

Abstract

Targeting zeros-emissions in transportation, future vehicles will be more energy-efficient via powertrain electrification. This PhD research aims to optimise an electrified off-highway vehicle to achieve the maximum energy efficiency by exploring new artificial intelligence algorithms. The modelling study of the vehicle system is firstly performed. Offline design optimisation and online optimum energy management control methodologies have been researched. New optimisation methods are proposed and compared with the benchmark methods. Hardware-in-the-Loop testing of the energy management controller has been carried out for validation of the control methods. This research delivers three original contributions:

1) Chaos-enhance accelerated particle swarm optimisation algorithm for offline design optimisation is proposed for the first time. This can achieve 200% higher reputation-index value compared to the particle swarm optimisation method.

2) Online swarm intelligent programming is developed as a new online optimisation method for model-based predictive control of the vehicle energy-flow. This method can save up to 17% energy over the rule-based strategy.

3) Multi-step reinforcement learning is researched for a new concept of ‘model-free’ predictive energy management with the capability of continuously online optimisation in real-world driving. It can further save at least 9% energy.

Type of Work:

Thesis (Doctorates > Ph.D.)

Award Type:

Doctorates > Ph.D.

Supervisor(s):