Title: Multi-objective optimisation of coordinated control strategy for reducing shift shock based on engine model with reduced-order

Authors: Xianhe Shang; Fujun Zhang; Zhenyu Zhang; Tao Cui

Addresses: School of Mechanical Engineering, Beijing Institute of Technology, Fundamental Science on Vehicular Power System Laboratory, No. 5, Zhongguancun South Street, Haidian District, Beijing, 10081, China ' School of Mechanical Engineering, Beijing Institute of Technology, Fundamental Science on Vehicular Power System Laboratory, No. 5, Zhongguancun South Street, Haidian District, Beijing, 10081, China ' School of Mechanical Engineering, Beijing Institute of Technology, Fundamental Science on Vehicular Power System Laboratory, No. 5, Zhongguancun South Street, Haidian District, Beijing, 10081, China ' School of Mechanical Engineering, Beijing Institute of Technology, Fundamental Science on Vehicular Power System Laboratory, No. 5, Zhongguancun South Street, Haidian District, Beijing, 10081, China

Abstract: To address the issue of increased shifting shocks caused by the limitations of the engine's full-speed regulation (FSR) characteristics during upshifting in heavy-duty vehicles, this paper proposes a coordinated control multi-objective optimisation strategy for reducing shifting shocks. This strategy takes into account the transient characteristics of the engine during the shifting process and uses long short-term memory (LSTM) neural networks to establish a reduced-order engine model. Based on the influence of the transient characteristics of the engine on shifting shock, a coordinated control scheme is formulated. To obtain the optimal solution for control parameters, a multi-objective optimisation was performed using the nondominated sorting genetic algorithm-II (NSGA-II) algorithm with the minimisation of root mean square of shifting shocks and friction work as optimisation objectives. Finally, the proposed coordinated control strategy was verified through simulation comparisons, demonstrating its superior control effectiveness in significantly reducing shifting shocks.

Keywords: LSTM neural network; NSGA-II; non-dominated sorting genetic algorithm-II; reduced-order model; shift shock; coordinated control.

DOI: 10.1504/IJHVS.2024.142259

International Journal of Heavy Vehicle Systems, 2024 Vol.31 No.6, pp.829 - 850

Received: 21 Oct 2023
Accepted: 03 Jan 2024
Published online: 16 Oct 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article