共查询到20条相似文献,搜索用时 453 毫秒
1.
为提高燃料电池混合动力汽车(FCHEV)燃料经济性以及维持蓄电池能量平衡,该文提出了基于等效因子的Q-learning算法的能量管理策略。构建等效耗氢量最小与维持蓄电池荷电状态(SOC)平衡的目标函数,建立FCHEV动力源能量流转化平衡模型,通过能量转化平衡机理得到耗氢量的等效因子;在城市循环+全球轻型汽车测试循环(UDDS+WLTC)工况下,对需求功率的转移概率矩阵进行求解,利用Q-learning算法离线优化燃料电池和蓄电池的输出功率;基于MATLAB/Simulink平台建立了前向仿真模型,进行整车性能的仿真试验。结果表明:在WLTC循环工况下,该策略的100 km等效耗氢量为0.730 kg,接近基于动态规则(DP)控制策略的耗氢量,且SOC保持在合理的范围内,验证了该策略的有效性;在西宁市实际工况下,验证了本文所提控制策略的适应性。 相似文献
2.
设计了一种具有实时控制能力的增程式电动汽车混合型能量管理策略。首先建立了面向能量管理策略优化的增程式电动汽车整车模型。根据能量管理策略特点,将优化目标设置为增程器系统燃油消耗及动力电池当前SOC值与目标值之间差值的总和。再采用动态规划算法求解增程式电动汽车在给定行驶工况下的能量管理优化问题,从而获得了增程器开启时刻与输出功率优化结果。但由于动态规划算法需要已知详细的工况信息,很难应用于实车实时控制,而且从动态规划优化结果中不易提取控制规则,因此利用BP神经网络算法对优化结果进行离线训练,建立了增程器输出功率与车辆行驶状态参数间的非线性映射关系,得到了具有实时控制能力的神经网络控制模型。在采用BP神经网络训练时,根据车辆各个状态参数在CAN总线中的传输精度,对神经网络输入层、输出层参数的精度进行了修正。仿真结果表明:神经网络模型能够获得类似动态规划的最优控制效果,能够控制动力电池SOC在目标值的3%误差带以内。采用NEDC工况对混合型能量管理策略进行了硬件在环仿真试验,试验结果表明:与实车采用的电能消耗-电能维持型控制策略相比,所提出的混合型能量管理策略使汽车的燃油经济性提高了9.5%。 相似文献
3.
针对THS-Ⅲ平台的插电式混合动力汽车提出一种基于深度强化学习的能量管理策略。首先,使用MATLAB/Simulink搭建车辆前向仿真模型;其次,建立车辆能量管理的马尔可夫过程和深度强化学习算法;最后,使用WLTC-Class3和ACC-60工况进行了仿真验证。结果表明,与基于规则的能量管理策略相比,基于深度强化学习的能量管理策略在WLTC-Class3工况下总花费节省16.51%,燃油消耗量下降15.56%,在ACC-60工况下总花费节省31.95%,燃油消耗量下降29.96%。 相似文献
4.
5.
为提高插电式混合动力汽车燃油经济性,设计了一种基于动态规划和径向基函数(RBF)神经网络的插电式混合动力汽车能量管理策略。首先,建立了插电式混合汽车数学模型;其次,以发动机油耗最小为目标函数,采用动态规划求解全局最优的离线优化结果;最后,采用RBF神经网络对离线最优控制结果进行学习,建立了发动机输出转矩与车辆状态参数之间的非线性映射关系,得到了基于动态规划和RBF神经网络的能量管理策略。仿真结果表明,文章所提策略油耗较之于电量消耗-维持策略降低了2.92%,验证了该策略的有效性。 相似文献
6.
7.
8.
9.
10.
文章针对一款串联插电式混合动力城市公交,提出一种可实时应用的模型预测控制(MPC)能量管理策略,以能耗最小为目标优化整车功率分配。首先,基于马尔科夫链根据历史车速和加速度建立单步和多步速度预测模型;从而进行预测时域内滚动优化,选择动态规划算法(DP)得到动力系统最优控制序列;最后对比了基于模型预测、动态规划和庞特里亚金极小值原理(PMP)的能量管理策略。结果表明,提出的模型预测控制(MPC)能达到与全局优化算法相近的控制效果且能应用于实时控制,是其他两种方法不具备的,体现出该策略的优越性。 相似文献
11.
Nadir Ouddah Lounis Adouane Rustem Abdrakhmanov 《International Journal of Automotive Technology》2018,19(3):571-584
This paper details the development of an energy management strategy (EMS) for real-time control of a multi hybrid plug-in electric bus. The energy management problem has been formulated as an optimal control problem in order to minimize the fuel consumption of the bus drivetrain for a typical day of operation. Considering the physical characteristics of the studied hybrid electric bus and its well-known daily tour, the Pontryagin’s minimum principle (PMP) is firstly used as the mean to obtain offline optimal EMS. Afterward, in order to adapt the proposed strategy for real-time implementation, the proposed control parameters are adapted online using feedback from the battery state of energy (SOE) which allows us to accurately control the battery SOE in the presence of wide range of uncertainties. The work proposed in this paper is conducted on a dedicated high-fidelity dynamical model of the hybrid bus, that was developed on MATLAB/TruckMaker software. The performance evaluation of the proposed strategy is carried out using a normalized driving cycles to represent different driving scenarios. Obtained results show that among the investigated methods, it is reasonable to conclude that the proposed adaptive online strategy based on PMP is the most suitable to design the targeted EMS. 相似文献
12.
13.
14.
15.
16.
C. H. Zheng G. Q. Xu Y. I. Park W. S. Lim S. W. Cha 《International Journal of Automotive Technology》2014,15(1):117-123
Pontryagin’s Minimum Principle (PMP) and Dynamic Programming (DP) are both from the optimal control theory and can both achieve optimal trajectories when they are applied to power management strategies of hybrid vehicles. However they have totally different control concepts. In order to select the superior one, the PMP-based and the DP-based power management strategies are introduced and compared for a fuel cell hybrid vehicle (FCHV) in this paper. The two power management strategies are applied to the FCHV in a computer simulation environment, and the simulation results from the two strategies are compared when the control variable for the PMP is fuel cell system (FCS) net power and for the DP is battery power. As a result, the superiority of the PMP-based power management strategy is proved. 相似文献
17.
基于对混合动力汽车能量管理策略优化的目的,建立了丰田Prius Plug-in混合动力汽车的MATLAB/Simulink数学模型,用数学公式描述了系统优化控制问题,采用粒子群优化算法对该包含众多约束条件的非线性优化问题进行了求解,利用PSAT专业软件对比分析了基本型优化控制算法、改进型优化控制算法和规则控制算法等的控制效果及燃油经济性。结果表明,经过优化后的Plug-in混合动力汽车在不牺牲汽车各项性能的前提下能提高动力系统工作效率。 相似文献
18.
19.
基于对混合动力汽车能量管理策略优化的目的,建立了丰田PnusPlug-in混合动力汽车的MATLAB/Simulink数学模型,用数学公式描述了系统优化控制问题,采用粒子群优化算法对该包含众多约束条件的非线性优化问题进行了求解,利用PSAT专业软件对比分析了基本型优化控制算法、改进型优化控制算法和规则控制算法等的控制效果及燃油经济性。结果表明,经过优化后的Plug-in混合动力汽车在不牺牲汽车各项性能的前提下能提高动力系统工作效率。 相似文献
20.
In this paper, a new methodology is presented for computing time-optimal obstacle avoidance maneuvers for ground vehicles. Usually, the problem of obstacle avoidance is addressed in two parts. In the first part a path is planned. In the second an appropriately designed vehicle controller tracks the desired path. In view of the fact that the main problem concerning emergency maneuvers remains the development of an optimal control for minimum time and maximum maneuverability — with respect to the slip risk due to saturation of the tire forces — the authors propose an alternative approach. Considering that the time optimal control according to Pontryagin’s Maximum Principle (PMP) is of bang-bang type the investigations concern the minimum order and magnitude bang-bang control for “feedforward” steering maneuvers with the target of minimizing the computation time and simplifying the algorithm. This is accomplished by keeping the basic PMP logic but transforming the computational algorithm from an exact to a least squares control problem. Furthermore, the paper addresses how to solve the problem of guiding the vehicle from a non rest to a rest position. Simulations of obstacle avoidance maneuvers illustrate the performance of the controller. 相似文献