基于分布参数系统和单粒子模型的锂离子电池工作状态实时监控模型

提出了一种基于分布参数系统偏微分方程描述和单粒子模型的锂离子电池工作状态实时监控建模方法,能够实时跟踪锂离子电池阳极的锂离子密度和残差变化,并通过仿真实验验证了本文模型对锂离子电池工作状态故障报警的有效性和精确性.     

基于扩展单粒子模型的锂离子电池参数识别策略

为了精确识别电动汽车锂离子动力电池的关键状态参数,基于多孔电极理论和浓度理论,建立了一种考虑液相动力学行为的锂离子电池扩展单粒子模型.相较于传统单粒子模型,该模型增加了对负电极表面固体电解质界面膜参数的描述,并考虑了温度和液相浓度变化对锂离子电池关键参数的耦合影响.基于所建立的扩展单粒子模型,提出一种简化的参数灵敏度分析方法和有效的锂电池参数识别策略,用以确定特定工况下的高灵敏度待识别参数,进而利用遗传算法实现参数的优化求解.最后,通过对比分析本文模型和传统单粒子模型的仿真输出电压和相同工况下电池的实验输出电压验证了提出模型和参数识别方法的有效性和可行性,为电池管理系统的健康状态估计提供了理论基础.     

包含液相扩散方程简化的锂离子电池电化学模型

锂离子电池的电化学模型对于电池特性分析和电池管理具有重要意义,但是准二维(P2D)模型复杂度太高,为了在保证模型精度的基础上尽量降低复杂度,本文提出了一种包含液相简化的P2D (LSP2D)模型,该模型首先基于电化学平均动力学将电池端电压化简成为仅耦合固相Li+浓度c_s和液相Li+浓度c_e的方程,然后进一步对表达c_s和c_e演化规律的偏微分方程进行抛物线近似化简,使得最终的模型由多项式组成.COMSOL仿真表明在放电倍率为1C时该模型与单粒子(SP)模型的估算精度和速度相当,但在放电倍率为3C时,该模型的估算时间比P2D模型减少了99.73%,与SP模型相当,估算精度相比SP模型有大幅度提升.     

基于电化学模型的锂离子电池多尺度建模及其简化方法

锂离子电池的精确建模和状态估计对于电动汽车电池管理系统非常重要,准二维(P2D)电化学模型由于计算复杂,难以直接应用于电池管理的参数在线估计和实时控制中.本文基于多孔电极理论和浓度理论,提出一种考虑锂离子液相动力学的简化准二维(SP2D)模型.忽略锂离子孔壁流量沿电极厚度方向的变化求解SP2D模型所描述的锂离子电池锂浓度分布,基于锂离子电池电化学平均动力学行为求解固相和液相电势变化,推导出电池电压计算的简化表达式;采用恒流、脉冲以及城市循环工况放电电流对比分析了严格P2D模型与SP2D模型的终端电压和浓度分布.结果表明:SP2D模型在保持较高计算精度的同时,可显著提高计算效率.     

基于电化学老化衰退模型的锂离子动力电池外特性

当前锂离子动力电池电化学模型存在模型复杂、建模难度大、计算效率低、老化评估效果差的问题,本文提出一种考虑电池衰退老化的机理模型(ADME).本文首先通过有限差分法对伪二维(P2D)电化学模型进行离散降阶处理,得到简化伪二维(SP2D)模型.在SP2D模型的基础上,基于阴阳两极发生的副反应导致的衰退老化现象,提出一种考虑电池衰退老化的机理模型.其次,使用多变量偏差补偿最小二乘法实现模型参数辨识.最后通过动力电池衰退老化性能循环实验,对比分析了恒流、脉冲工况下SP2D模型和ADME模型的终端电压输出.结果表明:ADME模型较为简单、计算效率和估算精度高,可以有效评估电池容量老化衰退,得到理想的锂离子动力电池外特性曲线.     

基于反激式直流变换器的电池均衡系统仿真研究

针对电动汽车锂离子电池组在工作过程中出现的能量不一致现象,设计了一种基于反激式直流变换器的主动均衡电路,包括电源转换、信号采集、驱动电路以及均衡器4个模块。研究了均衡系统的拓扑结构和工作原理以及控制策略,阐述了反激式直流变换器与锂离子电池的建模方法,针对电池组8种典型的非均衡工况,进行了混合策略下的5 kHZ高频仿真实验,实验用电池SOC从44.3%到55.7%均有涉及,仿真与实验结果误差均在10%以内,验证了PSIM仿真电路与同轴多绕组变压器与电池建模方法以及混合控制策略的正确性。     

一种求解锂离子电池单粒子模型液相扩散方程的新方法

电解液中的锂离子浓度表达是锂离子电池电化学模型求解的基本任务之一.为了平衡单粒子模型的液相动态性能和计算效率,假设反应仅发生在集电极和电解质界面上,为此,提出一种基于液相扩散方程无穷级数解析解的界面浓度求解新方法.在恒流工况下,利用数列单调收敛准则将解析解转化为一个收敛和函数.在动态工况下,将该解析解简化为输入与和函数的无限离散卷积.利用和函数随时间单调衰减并收敛至零的特性对其进行截断,从而得到有限离散卷积求解算法.对比专业有限元分析软件,该方法在恒流工况和动态工况下均能以较少的计算时间获得相当好的精度.而且,该方法仅有一个配置参数.因此,所提方法将有效减小应用于实时电池管理系统上的电化学模型计算负担.     

一种求解锂离子电池单粒子模型液相扩散方程的新方法

电解液中的锂离子浓度表达是锂离子电池电化学模型求解的基本任务之一.为了平衡单粒子模型的液相动态性能和计算效率,假设反应仅发生在集电极和电解质界面上,为此,提出一种基于液相扩散方程无穷级数解析解的界面浓度求解新方法.在恒流工况下,利用数列单调收敛准则将解析解转化为一个收敛和函数.在动态工况下,将该解析解简化为输入与和函数的无限离散卷积.利用和函数随时间单调衰减并收敛至零的特性对其进行截断,从而得到有限离散卷积求解算法.对比专业有限元分析软件,该方法在恒流工况和动态工况下均能以较少的计算时间获得相当好的精度.而且,该方法仅有一个配置参数.因此,所提方法将有效减小应用于实时电池管理系统上的电化学模型计算负担.     

光伏电池组件隐式、显式单二极管模型准确性对比研究

光伏电池组件非线性输出特性的物理建模及其优化参数的准确提取是光伏发电系统设计计算、性能评估及优化控制的重要前提.相对于传统的隐式单二极管模型,该文在光伏电池显式单二极管模型的基础上利用Lambert W函数推导了光伏组件的显式单二极管模型,提出一种基于重启边界约束Nelder-Mead单纯形算法的参数提取方法rbcNM,并利用两种典型光伏电池组件的实测数据对隐式、显式单二极管模型的准确性进行了对比测试和验证.结果表明:rbcNM算法可以快速准确的提取隐式、显式单二极管模型的优化参数,计算结果与实测数据具有很好的一致性,相对于已有文献在准确度上取得了大幅度的提升;显式单二极管模型的准确性显著高于隐式单二极管模型,对光伏电池组件的电流-电压和功率-电压特性曲线具有更高的拟合精度.     

基于三维激光扫描的物体重构建模

针对锂离子电池SOC(荷电状态)难以估算的问题,通过对电池建立等效的Thevenin电路模型,对不同时刻的SOC的模型参数进行拟合得到动态的模型参数,在Matlab中借助Simulink建立仿真模型,采用模块化结构,建立基于卡尔曼滤波算法的电池SOC估算系统;利用测得的电池电压电流,仿真系统可直接估算出实时的电池SOC,与实际的电池SOC对比,误差保持在2.5%以内,表明该方法可以有效的估计电池的SOC,对于锂离子电池在实际应用的容量估算有着重要意义。     

基于分布参数估计的锂离子电池故障预测建模

Lithium ion battery has typical character of distributed parameter system, and can be described precisely by partial differential equations and multi-physics theory because lithium ion battery is a complicated electrochemical energy storage system. A novel failure prediction modeling method of lithium ion battery based on distributed parameter estimation and single particle model is proposed in this work. Lithium ion concentration in the anode of lithium ion battery is an unmeasurable distributed variable. Failure prediction system can estimate lithium ion concentration online, track the failure residual which is the difference between the estimated value and the ideal value. The precaution signal will be triggered when the failure residual is beyond the predefined failure precaution threshold, and the failure countdown prediction module will be activated. The remaining time of the severe failure threshold can be estimated by the failure countdown prediction module according to the changing rate of the failure residual. A simulation example verifies that lithium ion concentration in the anode of lithium ion battery can be estimated exactly and effectively by the failure prediction model. The precaution signal can be triggered reliably, and the remaining time of the severe failure can be forecasted accurately by the failure countdown prediction module.     

Plasma kinetic control in a tokamak

Plasma kinetic (temperature and density) control is developed in terms of optimal control theory. This involved solving a distributed parameter control problem for which no previous general theory or techniques could be practically implemented. The methodology used involves reducing the one-dimensional particle and energy transport partial differential equations into a system of ordinary differential equations. The latter are linearized and put into standard control theory format. This technique can be generalized to any distributed parameter system whose variables can be modeled as simple analytic functions of the spatial coordinates. This distributed parameter control technique shows excellent control characteristics when applied to realistic plasma temperature and density profiles. Tests were made on temperatures and densities which had been perturbed about 10% below their desired value and profiles that were significantly more peaked than the required shape. Results were obtained for a simplified model problem with specific empirical transport coefficients and a one species plasma     

一种基于简化电化学模型的锂电池互联状态观测器

锂电池正、负极固相浓度分布以及荷电状态的精确估计对于开发锂电池工作状态的实时监控算法,进而构建高效、可靠的锂电池管理系统具有重要意义.本文基于多孔电极理论和浓度理论,提出基于扩展单粒子模型的锂电池关键内部参数识别的优化模型和方法;在该电化学模型简化的基础上,提出一种基于H_∞鲁棒控制理论框架的锂电池新型双向互联观测器,可同时实现对锂电池正、负电极浓度及荷电状态的估计,并通过对比分析不同工况下的仿真结果和实验数据,对所提出的互联观测器性能进行了系统验证.结果表明：所设计的互联观测器能够准确预测锂电池的输出电压和荷电状态,有效提高了锂电池系统模型的动态性能和鲁棒稳定性,为锂电池管理系统的开发奠定了理论基础.     

ELECTRIC CIRCUIT DESIGN AND TESTING OF INTEGRATED DISTRIBUTED STRUCTRONIC SYSTEMS

Modelling distributed parameter systems (DPS) by electric circuits and fabricating the complicated equivalent circuits to evaluate system responses poses many challenging research issues for many years. Electrical modelling and analysis of distributed sensing/control of smart structures and distributed structronic systems are even scarce. This paper is to present a technique to model distributed structronic control systems with electric circuits and to evaluate control behaviors with the fabricated equivalent circuits. Electrical analogies and analysis of distributed structronic systems is proposed and dynamics and control of beam/sensor/actuator systems are investigated. To determine the equivalent circuits and system parameters, higher order partial derivatives are simplified using the finite difference method; partial differential equations (PDE) are transformed to finite difference equations and further represented by electronic components and circuits. To provide better signal management and stability, active electrical circuit systems are designed and fabricated. Electrical signals from the distributed system circuits (i.e., soft and hard) are compared with results obtained by the classical theoretical, finite element, and experimental techniques.     

A single dual-mode gas sensor for early safety warning of Li-ion batteries: Micro-scale Li dendrite and electrolyte leakage

Li dendrites and electrolyte leakage are common causes of Li-ion battery failure. H₂, generated by Li dendrites, and electrolyte vapors have been regarded as gas markers of the early safety warning of Li-ion batteries. SnO₂-based gas sensors, widely used for a variety of applications, are promising for the early safety detection of Li-ion batteries, which are necessary and urgently required for the development of Li-ion battery systems. However, the traditional SnO₂ sensor, with a single signal, cannot demonstrate intelligent multi-gas recognition. Here, a single dual-mode (direct and alternating current modes) SnO₂ sensor demonstrates clear discrimination of electrolyte vapors and H₂, released in different states of Li-ion batteries, together with principal component analysis (PCA) analysis. This work provides insight into the intelligent technology of single gas sensors.     

Analysis of multichannel optical soliton collisions with asymmetric energy in strongly dispersion managed transmission networks considering higher-order effects

We study the interaction properties of optical soliton pulses propagating in a two-channel wavelength-division multiplexed strongly dispersion managed communication system. We analyze the effects of third order dispersion, Raman scattering and self-steepening using an ordinary differential equations model obtained by a variational method applied to the Generalized Nonlinear Schrödinger Equation. The validity of the model is assessed against the integration of the full nonlinear partial differential equations. The variational equations are initially solved for a single pulse in order to identify the launching parameters for each pulse in the first DM cell of the system. One pulse per wavelength is then injected in the transmission link. We then systematically study the evolution of multiplexed soliton trains and analyse the transmission system in terms of residual frequency shifts and interchannel interactions as the map strength is varied, focusing on the ratio of dissimilar peak powers in a broad range of dispersion difference values, concluding that the transmission characteristics improve by using specific values of unequal energies and considering higher-order correction terms. The work presented is an extension of previous analysis where isolated intra-channel two-pulse interactions had been addressed considering pulses with dissimilar energies.     

正冲击波后固壁表面颗粒的上升运动

为揭示固壁表面颗粒的冲击波夹卷本质，模拟垂直于固壁表面的正冲击波后单个颗粒的上升运动。假定颗粒初始时刻处于气载状态（因波的反射或碰撞离开壁面），受Saffman升力、气动阻力和重力作用。模型方程为波后定常气流边界层方程和颗粒运动常微分方程，分别用单参数法和四阶龙格库塔法求解。计算颗粒速度与轨迹表明：颗粒的冲击波卷扬动力，源于边界层内强剪切流提供的Saffman力；在所考察的冲击波强度和颗粒尺寸范围内，颗粒的上升高度不依赖于冲击波强度；上升高度按颗粒尺寸变化，即尺寸越大，高度越大。比较分析单颗粒和颗粒层的结果，认为实验中颗粒云内大尺寸颗粒少的原因是由于这些颗粒不易从颗粒层中逸出。部分计算结果与实验结果较为符合，验证了所建模型与假设的合理性。