采用数学模型法对磷酸铁锂电池进行非线性建模,优化了状态模型及观测模型.模型考虑了充放电倍率、温度、老化循环寿命等因素,对电池松弛效应及极化现象影响进行建模补偿,提高了电池建模的准确度,降低了不同条件下因电池模型造成电池荷电状态(SOC)估算的误差影响.在电池模型参数辨识基础上,提出采样自适应Sigma卡尔曼算法构建SOC估算模型,按照非线性模型对状态变量的分布构建Sigma采样序列,采用模型输出残差更新噪声协方差,赋予Sigma采样序列最优估计及噪声的权值,并实现误差量的实时更新,降低计算复杂度.通过持续大电流、间断电流、变电流放电及充电实验条件下的SOC估算对比实验,验证了自适应Sigma卡尔曼算法快速收敛性,数学描述更准确,具备较高的SOC的观测准确度.%The nonlinear model was applied to describe the lithium iron phosphate battery by mathematical model method,and the status model and observation model were optimized.Take into consideration the influences of charge-discharge rate,temperature variation and aging cycle life,the status model was improved.The observation model was also compensated for battery relaxation and polarization effect.Thus,the battery modeling accuracy was enhanced,and state of charge (SOC)estimation error caused by battery model was reduced under different conditions.Then,on the basis of the parameters identification for battery model,an improved adaptive sigma Kalman filter algorithm was proposed to construct the SOC estimation model.According to state variables distribution of the nonlinear model,the sigma sample sequence was built.Each model residual error covariance of output was used to update the covariance of the noise in real time.The optimal estimate sigma sampling sequence and the weights of noise were also reassigned by real-time updates with low computational complexity.The experiments were carried out to estimate the properties by charge,continuous-current discharge,backlash-current discharge and varying current discharge mode.The results verify the rapid convergence and more accurate mathematical description.It is shown that the accuracy of SOC estimation is improved using proposed model and algorithm.
展开▼
