A polynomial model, time origin shifting model(TOSM, is used to describe the trajectory of a moving target .Based on TOSM, a recursive laeast squares(RLS) algorithm with varied forgetting factor is derived for tracki...A polynomial model, time origin shifting model(TOSM, is used to describe the trajectory of a moving target .Based on TOSM, a recursive laeast squares(RLS) algorithm with varied forgetting factor is derived for tracking of a non-maneuvering target. In order to apply this algorithm to maneuvering targets tracking ,a tracking signal is performed on-line to determine what kind of TOSm will be in effect to track a target with different dynamics. An effective multiple model least squares filtering and forecasting method dadpted to real tracking of a maneuvering target is formulated. The algorithm is computationally more effcient than Kalman filter and the percentage improvement from simulations show both of them are considerably alike to some extent.展开更多
Li-ion batteries are widely used in electric vehicles(EVs).However,the accuracy of online SOC estimation is still challenging due to the time-varying parameters in batteries.This paper proposes a decoupling multiple f...Li-ion batteries are widely used in electric vehicles(EVs).However,the accuracy of online SOC estimation is still challenging due to the time-varying parameters in batteries.This paper proposes a decoupling multiple forgetting factors recursive least squares method(DMFFRLS)for EV battery parameter identification.The errors caused by the different parameters are separated and each parameter is tracked independently taking into account the different physical characteristics of the battery parameters.The Thevenin equivalent circuit model(ECM)is employed considering the complexity of battery management system(BMS)on the basis of comparative analysis of several common battery ECMs.In addition,decoupling multiple forgetting factors are used to update the covariance due to different degrees of error of each parameter in the identification process.Numerous experiments are employed to verify the proposed DMFFRLS method.The parameters for commonly used LiFePO4(LFP),Li(NiCoMn)O2(NCM)battery cells and battery packs are identified based on the proposed DMFFRLS method and three conventional methods.The experimental results show that the error of the DMFFRLS method is less than 15 mV,which is significantly lower than the conventional methods.The proposed DMFFRLS shows good performance for parameter identification on different kind of batteries,and provides a basis for state of charge(SOC)estimation and BMS design of EVs.展开更多
针对多种电机采用的控制器模型也存在不同,电机参数存在多样性,并且针对多种工况下,电机在工作过程中存在参数变化、负载扰动等情况,导致电机参数辨识精度不高、电机-控制器模型失配等问题,提出了一种改进递归最小二乘法(recursive leas...针对多种电机采用的控制器模型也存在不同,电机参数存在多样性,并且针对多种工况下,电机在工作过程中存在参数变化、负载扰动等情况,导致电机参数辨识精度不高、电机-控制器模型失配等问题,提出了一种改进递归最小二乘法(recursive least squares,RLS)算法进行多种工况下多电机参数失配诊断。针对传统的递归最小二乘法在进行在线电机参数辨识时,容易固遗忘因子影响,存在跟随速度慢、抗干扰性差等问题,在原始递归最小二乘法基础上引入了随系统工况变化而变化的“变遗忘因子”,提高电机参数的跟踪速度和抗负载扰动能力;为验证改进后的递归最小二乘法是否具有可靠性、鲁棒性和泛化性,分别设置了5种假设工况,并进行多组实验对比,通过分析电机速度响应、d-q轴电流以及量化分析转矩跟踪和R_(s)参数辨识精度等,验证改进后的算法具有较强的鲁棒性和泛化性;并通过分析性能指标数据,主要包括平均速度、平均q轴电流等,得出改进后算法分析数据的有效性。展开更多
荷电状态(state of charge,SOC)的准确估计对延长电池寿命、减少事故发生至关重要。针对锂电池系统存在建模误差及宽温度范围下传统方法适应性差的问题,设计一种自适应增益滑模观测器(adaptive gain sliding mode observer,AGSMO)以提...荷电状态(state of charge,SOC)的准确估计对延长电池寿命、减少事故发生至关重要。针对锂电池系统存在建模误差及宽温度范围下传统方法适应性差的问题,设计一种自适应增益滑模观测器(adaptive gain sliding mode observer,AGSMO)以提高宽温域SOC估计精度。采用二阶RC等效电路模型构造适用于AGSMO的状态方程,并结合遗忘因子最小二乘法(forgetting factor recursive least square,FFRLS)完成模型参数辨识。利用等效控制思想构建状态误差的等效表达式,基于此设计滑模观测器,同时采用自适应增益提高收敛速度并抑制抖振。结合案例应用仿真,结果表明:AGSMO在美国联邦城市运行工况FUDS和高加速循环工况US06的不同初值下均可实现SOC的准确估计,并通过上述两种工况验证宽温域环境下AGSMO相较于滑模观测器(sliding mode observer,SMO)、扩展卡尔曼滤波(extended Kalman filter,EKF)具有更好的估计精度及收敛速度,均方根误差不超过0.68%,且在温域两端呈现强鲁棒性。展开更多
为了实现永磁同步直线电机PMSLM(permanent magnet synchronous linear motor)高精度的多电气参数在线辨识,提出了一种基于双模型的递推最小二乘电气参数在线辨识算法。首先,根据电机的dq轴电压方程分别建立了辨识定子电阻、永磁体磁链...为了实现永磁同步直线电机PMSLM(permanent magnet synchronous linear motor)高精度的多电气参数在线辨识,提出了一种基于双模型的递推最小二乘电气参数在线辨识算法。首先,根据电机的dq轴电压方程分别建立了辨识定子电阻、永磁体磁链的模型1和辨识q轴电感、d轴电感的模型2,并将2个辨识模型循环结合。其次,基于上述双模型结构,采用递推最小二乘算法实现电气参数在线辨识,并针对PMSLM运行时存在大量动态过程的特性,提出一种具有饱和特性的分段变遗忘因子;然后,对功率开关非理想因素导致的误差电压进行补偿,进一步提高了辨识的精准度;最后,仿真和实验结果证明了该辨识算法的有效性,且具有收敛速度快、辨识结果精度高、多工况适用等优点。展开更多
文摘A polynomial model, time origin shifting model(TOSM, is used to describe the trajectory of a moving target .Based on TOSM, a recursive laeast squares(RLS) algorithm with varied forgetting factor is derived for tracking of a non-maneuvering target. In order to apply this algorithm to maneuvering targets tracking ,a tracking signal is performed on-line to determine what kind of TOSm will be in effect to track a target with different dynamics. An effective multiple model least squares filtering and forecasting method dadpted to real tracking of a maneuvering target is formulated. The algorithm is computationally more effcient than Kalman filter and the percentage improvement from simulations show both of them are considerably alike to some extent.
基金This work was supported by Science and Technology Project of State Grid Corporation of China(5202011600U5).
文摘Li-ion batteries are widely used in electric vehicles(EVs).However,the accuracy of online SOC estimation is still challenging due to the time-varying parameters in batteries.This paper proposes a decoupling multiple forgetting factors recursive least squares method(DMFFRLS)for EV battery parameter identification.The errors caused by the different parameters are separated and each parameter is tracked independently taking into account the different physical characteristics of the battery parameters.The Thevenin equivalent circuit model(ECM)is employed considering the complexity of battery management system(BMS)on the basis of comparative analysis of several common battery ECMs.In addition,decoupling multiple forgetting factors are used to update the covariance due to different degrees of error of each parameter in the identification process.Numerous experiments are employed to verify the proposed DMFFRLS method.The parameters for commonly used LiFePO4(LFP),Li(NiCoMn)O2(NCM)battery cells and battery packs are identified based on the proposed DMFFRLS method and three conventional methods.The experimental results show that the error of the DMFFRLS method is less than 15 mV,which is significantly lower than the conventional methods.The proposed DMFFRLS shows good performance for parameter identification on different kind of batteries,and provides a basis for state of charge(SOC)estimation and BMS design of EVs.
文摘针对多种电机采用的控制器模型也存在不同,电机参数存在多样性,并且针对多种工况下,电机在工作过程中存在参数变化、负载扰动等情况,导致电机参数辨识精度不高、电机-控制器模型失配等问题,提出了一种改进递归最小二乘法(recursive least squares,RLS)算法进行多种工况下多电机参数失配诊断。针对传统的递归最小二乘法在进行在线电机参数辨识时,容易固遗忘因子影响,存在跟随速度慢、抗干扰性差等问题,在原始递归最小二乘法基础上引入了随系统工况变化而变化的“变遗忘因子”,提高电机参数的跟踪速度和抗负载扰动能力;为验证改进后的递归最小二乘法是否具有可靠性、鲁棒性和泛化性,分别设置了5种假设工况,并进行多组实验对比,通过分析电机速度响应、d-q轴电流以及量化分析转矩跟踪和R_(s)参数辨识精度等,验证改进后的算法具有较强的鲁棒性和泛化性;并通过分析性能指标数据,主要包括平均速度、平均q轴电流等,得出改进后算法分析数据的有效性。
文摘荷电状态(state of charge,SOC)的准确估计对延长电池寿命、减少事故发生至关重要。针对锂电池系统存在建模误差及宽温度范围下传统方法适应性差的问题,设计一种自适应增益滑模观测器(adaptive gain sliding mode observer,AGSMO)以提高宽温域SOC估计精度。采用二阶RC等效电路模型构造适用于AGSMO的状态方程,并结合遗忘因子最小二乘法(forgetting factor recursive least square,FFRLS)完成模型参数辨识。利用等效控制思想构建状态误差的等效表达式,基于此设计滑模观测器,同时采用自适应增益提高收敛速度并抑制抖振。结合案例应用仿真,结果表明:AGSMO在美国联邦城市运行工况FUDS和高加速循环工况US06的不同初值下均可实现SOC的准确估计,并通过上述两种工况验证宽温域环境下AGSMO相较于滑模观测器(sliding mode observer,SMO)、扩展卡尔曼滤波(extended Kalman filter,EKF)具有更好的估计精度及收敛速度,均方根误差不超过0.68%,且在温域两端呈现强鲁棒性。
文摘为了实现永磁同步直线电机PMSLM(permanent magnet synchronous linear motor)高精度的多电气参数在线辨识,提出了一种基于双模型的递推最小二乘电气参数在线辨识算法。首先,根据电机的dq轴电压方程分别建立了辨识定子电阻、永磁体磁链的模型1和辨识q轴电感、d轴电感的模型2,并将2个辨识模型循环结合。其次,基于上述双模型结构,采用递推最小二乘算法实现电气参数在线辨识,并针对PMSLM运行时存在大量动态过程的特性,提出一种具有饱和特性的分段变遗忘因子;然后,对功率开关非理想因素导致的误差电压进行补偿,进一步提高了辨识的精准度;最后,仿真和实验结果证明了该辨识算法的有效性,且具有收敛速度快、辨识结果精度高、多工况适用等优点。
