基于混合储能系统与工况识别的电动拖拉机多目标能效管理优化

Abstract The promotion of electric tractors faces significant challenges, including adapting powertrain systems to diverse operational conditions and optimizing energy efficiency and battery lifespan. This paper presents a hybrid energy storage system (HESS) architecture for electric tractors. And a multi-objective energy-efficient management strategy (EMS) based on plowing operation scenario recognition is proposed. The strategy involves developing an electric tractor model and a plowing operating condition (POC) cycle using real-world plowing data. Offline classification is performed using K-means clustering and Principal Component Analysis (PCA), while a Multilayer Perceptron Neural Network (MLPNN) is employed for online real-time scenario recognition. Additionally, a Multi-Strategy Improved Black-winged Kite Algorithm (MSIBKA) is developed to efficiently derive adaptive power allocation trajectories. Simulation and Hardware-in-the-Loop (HIL) experiments demonstrate that the proposed strategy effectively extends the lifespan of the HESS, smooths battery output, and reduces operating costs. Specifically, the supercapacitor supplies over 65 % of the peak power demand, reducing the battery C-rate by more than 10 %. Furthermore, the proposed system increases the state of charge (SOC) of the battery by at least 5 %, while reducing both operational costs and battery degradation costs by over 33.3 %. These results indicate that the proposed system and strategy provide substantial benefits in extending battery lifespan and enhancing energy efficiency.