Multi-objective energy management for Atkinson cycle engine and series hybrid electric vehicle based on evolutionary NSGA-II algorithm using digital twins. (15th February 2021)

• Record Type: Journal Article
• Authors: Li, Yangyang; Wang, Shuqian; Duan, Xiongbo; Liu, Shujing; Liu, Jingping; Hu, Song
• Is Part Of: Energy conversion and management. Volume 230(2021)
• Publication Date: 2021-02-15
• Subjects: Atkinson engine -- Series hybrid electric vehicle -- Evolutionary non-dominated sorting genetic algorithm -- Exhaust gas recirculation rate -- Fuel consumption -- NO emissions; ICE internal combustion engine -- HCCI homogeneous charge compression ignition -- VVA variable valve angle -- EGR exhaust gas recirculation -- GDI gasoline direct injection -- CNG compressed natural gas -- HEV hybrid electric vehicle -- VVT variable valve timing -- NO nitric oxide -- VOAC variable Otto-Atkinson cycle -- HAVT hydraulic actuated valvetrain -- DI direct injection -- HC Hydrocarbon -- CO carbon monoxide -- CO2 carbon dioxide -- LIVC late intake -valve closure -- SA spark angle -- EVO exhaust valve open -- AFR air–fuel-ratio -- ANN artificial neural network -- GA genetic algorithm -- SVM support vector machine -- GRNN generalized regression neural network -- PSO particle swarm optimization -- NSGA-II non-dominated sorting genetic algorithm-II -- HPDI high pressure direct injection -- EER effective expansion ratio -- BSFC brake specific fuel consumption -- MOEO multi objective evolutionary optimization -- VVT-I intake variable valve timing -- VVT-e exhaust variable valve timing -- EEE effective expansion efficiency -- SOC start of combustion -- EOC end of combustion -- BDC bottom dead center -- CR compression ratio -- KI knock index -- RGF residual gas coefficient -- PMEP pumping mean effective pressure -- NEDC new European driving cycle -- CFD computational fluid dynamics; Direct energy conversion -- Periodicals; Energy storage -- Periodicals; Energy transfer -- Periodicals; Énergie -- Conversion directe -- Périodiques; Direct energy conversion; Periodicals; 621.3105
• DOI: 10.1016/j.enconman.2020.113788
• Languages: English
• ISSNs: 0196-8904

Abstract

Highlights: High-accuracy combined simulation–optimization platform for the vehicle is developed. NSGA-II is put forward, clarified and applied for optimization of full engine MAPs. Reduction rates of fuel consumption and NO are decreased by up to 12.48% and 92.64%. Cumulative fuel consumption and NO of optimized vehicle reduced by 4.58% and 46.1%. Abstract: In order to develop higher performance Atkinson cycle gasoline engine and explore its fuel-saving potential on series hybrid electric vehicles, this study is pioneered in digital twins by GT-Power software, MATLAB/Simulink software and multi objective evolutionary optimization using evolutionary non-dominated sorting genetic algorithm. In the first stage, an experimental investigation is carried out and a corresponding 1-D GT-Power simulation model is developed and validated by the experimental data for an Otto cycle engine and then modified into the Atkinson cycle engine. In the second stage, the digital twins engine model takes the spark timing, exhaust gas recirculation rate, intake variable valve timing, exhaust variable valve timing as well as lambda as the inputs of the simulation optimization platform for the Atkinson cycle engine. The optimum values of aforementioned input parameters are identified by the evolutionary non-dominated sorting genetic algorithm to minimize the brake specific fuel consumption and nitric oxide under different speeds and loads, the reduction rates of fuel consumption and nitric oxide are decreased by up to 12.48% and 92.64%, respectively. In the third stage, the optimized performance MAPs are implemented in the series hybrid electric vehicle with the Atkinson cycle engine, the results show that the cumulative fuel consumption and nitric oxide volume fraction of the optimized vehicle under new European driving cycle reduced by 4.58% and 46.1%, respectively. It is concluded that the proposed evolutionary non-dominated sorting genetic algorithm method can identify the optimum conditions of vehicle well and improve its fuel economy as well as emission. Furthermore, the combined simulation platform for both engine and vehicle can be applied to evaluate and optimize the energy distribution and performance of vehicles with different technologies or strategies in the future. Besides, all these will provide theoretical basis and digital model support for the development of efficient and energy-saving new energy vehicles.