A P2-Configuration PHEV Energy Management Strategy Integrating a Novel Frequency-Reduction Algorithm for ICE Start–Stop Events

Aimed at addressing the issue of frequent short-duration ICE start–stop events in P2-configuration plug-in hybrid electric vehicles (PHEVs) employing conventional instantaneous optimization-based Equivalent Consumption Minimization Strategy (ECMS)—and the resulting deterioration of vehicle smoothness, NVH performance, fuel economy, and emission performance—this paper proposes a novel energy management strategy, designated ECMS-ISS, which integrates instantaneous optimization with an engine unnecessary start suppression algorithm. A multilayer perceptron (MLP) neural network is first constructed as an online identifier to recognize high-frequency intervals of frequent start–stop events in real time. A dedicated penalty function is then embedded within these identified intervals, with the penalty intensity adaptively adjusted according to the accumulated count of short-duration start–stop events, enabling zoned and targeted intervention without affecting engine torque output during normal operating intervals. Simulation results under NEDC and WLTC driving cycles demonstrate that, compared with the conventional A-ECMS, ECMS-ISS reduces engine start–stop events by 35.48% and 32.31%, respectively, and reduces comprehensive fuel consumption by 2.13% and 5.40%, while significantly decreasing CO, NOx, and HC emissions. Compared with RB-EMS, ECMS-ISS also exhibits superior fuel economy and emission reductions, with the final SOC maintained within a reasonable range throughout. The proposed strategy demonstrates distinct advantages in reconciling multiple objectives, including start–stop rationality, fuel economy, emission performance, and battery health, thereby providing a practical and adaptive solution to the frequent engine start–stop problem in P2-configuration PHEVs.