摘要
为解决一般电池模型对于不同健康状况电池泛化性能较差的问题,利用机器学习中最小二乘支持向量机(LSSVM)的原理,通过提取锂电池运行过程中的外部参数构建LSSVM模型。引入粒子群优化算法(PSO)来提高训练的效率与模型准确性^([1])。通过恒流放电实验比较了几种核函数和几种优化算法的估计效果,验证了PSO-LSSVM模型在复杂运行状况下电池荷电状态(SOC)估计的有效性。并与其他方法进行比较,进一步验证方案的优越性。该方法为新能源动力汽车的进一步发展提供了有效的技术支持。
In order to solve the problem that the general battery model has poor generalization performance for batteries in different health conditions,the LSSVM model is built by extracting the external parameters during the operation of lithium battery by using the principle of least square support vector machine(LSSVM)in machine learning.The particle swarm optimization algorithm(PSO)is introduced to improve the efficiency of training and model accuracy.By comparing the estimation results of several kernel functions and several optimization algorithms,the PSO-LSSVM model is verified to be effective in estimating battery state-of-charge(SOC)under complex operating conditions.Compared with other methods,the superiority of the scheme is further verified.This method provides effective technical support for the further development of new energy powered vehicles.
作者
戴庚
耿诗尧
Dai Geng;Geng Shiyao(Ocean University of China,Qingdao Shandong 266100,China)
出处
《信息与电脑》
2018年第2期31-32,共2页
Information & Computer
基金
锂离子电池建模及SOC精确估算系统设计(项目编号:201717900261)
关键词
LSSVM荷电状态
核函数
粒子群优化
LSSVM state of charge
kernel function
particle swarm optimization
