Model-based design of an energy management strategy for a hybrid electric bus

  • Dan Smith

Student thesis: Doctoral ThesisDoctor of Philosophy


This thesis provides an in-depth exploration of the state of the art in hybrid electric vehicle energy management. A broad literature review was conducted which revealed that most research has focused on strategies that can perform real-time optimisation, however, the real-world implementation of these systems is generally neglected. For vehicle applications, the control methods must be computationally efficient due to the limited processing power and memory available. In light of these findings, a control system optimised for real-world implementation on a hybrid electric bus was investigated. Detailed component models were generated using a combination of first principles and system identification. A novel artificial neural network-based fuel consumption model was created, enabling highly accurate energy consumption estimations. Using real-world test data, individual component models were validated and later incorporated into a complete model of the vehicle. A fuzzy system that controlled the electric motor torque based on the engine and drivetrain efficiency was developed and optimised using the vehicle model and a genetic algorithm. The optimised fuzzy system demonstrated an energy consumption reduction of approximately 10 % relative to a traditional deterministic rule-based strategy. Real-world tests were conducted that demonstrated a real-world energy consumption reduction of up to 20 % relative to the existing strategy present on the target vehicle. These results demonstrate that the proposed system is suitable for real-world implementation while also achieving improved performance relative to conventional methods.
Date of AwardJul 2022
Original languageEnglish
Awarding Institution
  • Queen's University Belfast
SponsorsNorthern Ireland Department for the Economy & Wrightbus Ltd
SupervisorWasif Naeem (Supervisor) & Roy Douglas (Supervisor)


  • HEV
  • EMS
  • hybrid vehicle
  • energy management strategy
  • fuzzy Logic
  • neural network
  • bus
  • model based design
  • genetic agorithm
  • particle swarm optimisation

Cite this