Research on fuel injection quantity fluctuation characteristics and optimization improvement of dual-fuel engines

The fluctuation of fuel injection quantity directly affects the power and economy of dual-fuel engines. Identifying key factors causing these fluctuations and improving injection stability are essential for better economy and emissions performance . This paper constructs and validates a high-pressure fuel injection system simulation model using a test platform. It examines the relationship between injection quantity fluctuations and needle action, finding that low injection pressure and narrow pulse width prevent rapid and full needle opening, leading to higher injection quantity fluctuation rate ( IQFR ). Specifically, low injection pressure reduces needle acceleration, while narrow pulse width limits needle ascent time. A data-driven model of IQFR is established using generalized regression neural network (GRNN) in combination with grey wolf optimizer (GWO). An optimization and improvement study is then carried out with the objective of reducing the fluctuation rate of the injection quantity. The results demonstrate that the maximum lift of the needle is diminished from 0.25 mm to 0.16 mm without a concomitant reduction in the maximum injection rate . The IQFR is diminished by a maximum of 13.3 % across the full range of operating conditions, achieving an enhancement in injection stability under conditions of low injection pressure and narrow injection pulse width. • This paper examines the relationship between injection quantity fluctuations and needle action. • The strong coupling between fuel pressure fluctuation and needle response is elucidated. • By employing the MOGWO algorithm, the structure of the injector is optimized without sacrificing the maximum injection rate. • The fluctuation of fuel injection for the improved injector is reduced across all operating conditions. • The method and model presented in this paper can be applied to similar fuel injection systems after adaptive adjustment.