|国家科技期刊平台
首页|期刊导航|农机化研究|基于需求功率预测的电动拖拉机能量管理策略

基于需求功率预测的电动拖拉机能量管理策略OA

Energy Management Strategy of Electric Tractor Based on Power Demand Prediction

中文摘要英文摘要

针对电动拖拉机在犁耕工况下电机需求电流波动比较大的特点,为了改善动力电池的输出电流过高或过低及电动拖拉机犁耕持续作业时间短的现象,利用超级电容高功率密度的特点,设计了一种锂电池+超级电容结构的双电源电动拖拉机,并建立了 Amesim/Simulink 联合仿真模型.以模型预测控制作为双电源系统的能量管理方法,基于长短期记忆神经网络建立电动拖拉机犁耕工况下的需求功率预测模型,使用动态规划算法求解最佳的锂电池输出电流.仿真结果表明:相比于模糊控制策略,基于模型预测控制策略有效降低了锂电池大电流放电的频率且峰值电流降低了 40%,有效提高了锂电池的使用寿命;超级电容的SOC 保持在比较高的范围内,且电动拖拉机在犁耕工况下的单位里程能量消耗降低了 2.17%,实现了双电源电流分配最优,提高了电动拖拉机的动力性和经济性.

In order to improve the phenomenon that the output current of power battery is too high or too low and the con-tinuous operation time of electric tractor is short,a dual power supply electric tractor with lithium battery as the main en-ergy and super capacitor as the auxiliary energy is designed by using the characteristics that supercapacitors have high power density,the AMESim/Simulink joint simulation model is established.In this paper,model predictive control is used as the energy management method of dual power supply system.Based on long-term and short-term memory neural net-work,the power demand prediction model of electric tractor under ploughing condition is established,and the dynamic programming algorithm is used to solve the optimal output current of lithium battery.The simulation results show that com-pared with the fuzzy control strategy,the model-based predictive control strategy effectively reduces the high current dis-charge frequency of lithium battery,reduces the peak current by 40%,and effectively improves the service life of lithium battery;The SOC of the super capacitor is kept in a relatively high range,and the energy consumption per unit mileage of the electric tractor under the ploughing condition is reduced by 2.17%,which realizes the optimal distribution of dual power supply current and improves the power performance and economy of the electric tractor.

盛志鹏;夏长高;孙闫;韩江义

江苏大学 汽车与交通工程学院,江苏 镇江 212013

农业工程

纯电动拖拉机双电源模型预测控制长短期记忆神经网络能量管理

pure electric tractordual power supplymodel predictive controllong short-term memory neural networkenergy management

《农机化研究》 2024 (005)

216-221 / 6

苏北科技专项-先导性项目(SZ-YC202165);江苏省重点研发计划项目(BE2018343-1)

评论