【作者】 刘辉；

【导师】 凌睿；

【作者基本信息】 重庆大学 ， 工程硕士（控制工程）（专业学位）， 2018， 硕士

【摘要】 为了延长锂离子电池组整体的使用寿命,减小过充过放电问题,在锂离子电池组使用的过程中需要对电池组进行均衡管理。基于双向Cuk变换器的总线式均衡电路,能够实现一对多,多对一和多对多的能量传输,均衡速率高,能量损耗少,且易于模块化,是近年来提出的新型均衡电路结构。总线式均衡电路具有多参数耦合、非线性等特点,使得均衡系统的建模控制难度较大。论文介绍了电池均衡技术研究的意义以及国内外的研究现状,对比分析了现有的均衡电路结构和均衡控制算法,提出一种基于模型预测控制(MPC)的双闭环控制策略对总线式均衡系统进行控制,外环控制器采用MPC,求解各个单节电池的均衡电流设定值,内环控制器使用PI控制器,实现均衡电流和总线电压的快速有效控制。主要工作如下:(1)研究了总线式均衡电路的工作原理,结合锂离子电池的充放电特性以及变换器的输入输出特性,建立了总线式均衡电路中能量转移的状态空间模型,给出了模型中变量的约束条件。(2)基于建立的状态空间模型,得到总线式均衡系统的预测状态。根据预测状态建立描述均衡过程损耗的目标函数,将求解最优均衡电流设定值的预测问题转化为非线性规划问题,并将均衡速度考虑进目标函数,通过近似分析进一步将均衡电流设定值的求解转化为线性规划问题。(3)研究了总线式均衡电路中总线电压以及均衡电流控制的原理。分别针对总线电压控制以及均衡电流控制,建立了均衡器的小信号模型。依据小信号模型分别设计了总线电压以及均衡电流PI控制器,实现了总线电压以及电流的快速无静差控制。分析了总线电压控制与均衡电流控制的相互影响。(4)设计了基于Matlab GUI的均衡系统样机。样机采用上下位机的结构形式,上位机实现MPC运算,下位机实现均衡电流和总线电压的实时控制。使用均衡系统样机分别进行了总线电压控制实验,均衡电流控制实验,能量转移控制实验,系统优化控制实验。实验结果验证了论文所设计的双闭环控制策略能够对基于双向Cuk变换器的总线式均衡电路进行快速有效的均衡控制。更多还原

【Abstract】 Lithium-ion batteries are among the most promising batteries.However,due to manufacturing variance,state of charge(SOC)imbalance is common in battery string,which can be called battery inconsistency and inevitably results in overcharge or over-discharge,even irreversible damage to cells collected in series.Battery management system with equalizing techniques makes great sense to improve battery uniformity.Energy bus-based approaches have high efficiency and fast equalization speed,all equalizers can work simultaneously to meet fast energy dispatch through energy bus.Meanwhile,the efficiency of equalization can be greatly improved.However,it is hard to be controlled due to its high order,non-linear,multi-parameter coupling and large time delay.By comparing existing equalization techniques and control strategies,model predictive control(MPC)is adopted to provide equalization currents reference of each bidirectional Cuk equalizer,and lag Compensator(PI controller)is implemented to regulate the real-time equalization currents and bus voltage.The main work of this paper is as follows:(1)The operation principle of energy bus-based equalization circuit is analyzed,according to which the state space model with constraints that describe the transfer of energy in the bus-based equalization circuit is derived.(2)Based on the state space model,predicted states can be obtained.Cost function that describes the lost during equalization is set up.Then the prediction of the equalization current references can be solved by non-linear programming.Furthermore,equalization speed is considered into cost function,current references can be obtained by linear programming by approximating equalizer efficiency to a constant.(3)The parallel operation of DC-DC converter is analyzed to come up with the principle of equalization current control and bus voltage control.Small signal model of Cuk converter is built.Equalization current and bus voltage PI controller are derived respectively based on the model to achieve a rapid response without steady-state error.Mutual interference between two control loops is given out.(4)Equalization system is designed to validate proposed control strategy.Host computer is implemented by Matlab GUI and executes the algorithm of MPC.Slave computer is implemented by STM32 and regulate equalization currents and bus voltage in real time.Current control experiment,bus voltage control experiment,energy transferred experiment and 5 cells equalization experiment are carried out and the results show that proposed control strategy can equalize the series connected batteries effectively and rapidly.更多还原