粒子群优化BP神经网络光伏锂电池充电系统  

Design of lithium battery charging control system using solar photovoltaic cells based on particle swarm optimization BP neural network

作　　者：陶佰睿[1,2] 葛思绮 郭琴[1] 苗凤娟[1] 张成军[2] 

机构地区：[1]齐齐哈尔大学通信与电子工程学院,黑龙江齐齐哈尔161006 [2]齐齐哈尔大学计算中心,黑龙江齐齐哈尔161006

出　　处：《电源技术》2016年第12期2411-2414,共4页Chinese Journal of Power Sources

基　　金：国家自然科学基金(61204127);黑龙江省教育科学"十二五"规划课题(GBC1214089);黑龙江省高校科技成果产业化前期研发培育项目(1254CGZH04);黑龙江省教育厅科学技术研究项目(12541899;12531774)

摘　　要：通过对太阳能光伏电池最大功率点的跟踪和锂电池的充电控制两部分实现了对锂电池的快速充电。首先利用SIMULINK的s-function的自定义模块和粒子群优化BP神经网络算法实现了对在辐射度、电池板温度、环境温度和风速共4种工作环境下的光伏电池进行最大功率点跟踪和建模,然后通过单片机编程对开关管进行PWM控制,完成充电电路的智能控制。之后,结合MATLAB和SIMULINK完成粒子群优化BP神经网络并在PROTEUS平台上完成充电电路的仿真,最后程序下载到硬件中并通过调试完成对锂电池充电过程的检测。结果表明,粒子群优化后的BP神经网络光伏锂电池充电控制系统具有较高的充电速度和充电效率。A lithium batteries fast charging control system using solar photovoltaic cells was designed by maximum power point Tracking for solar cells and Particle Swarm Optimization BP Neural Network controlling. Firstly, the SIMULINK S-function module and the BP neural network of particle swarm optimization algorithm were employed to achieve four work environment parameters such as radiometry, battery plate temperature, ambient temperature and wind speed for modeling and maximum power point tracking. Then, the intelligent control of the charging circuit was designed by PWM controlling switch transistor with programming SCM. Next, the PSO BP neural network system simulation was implemented by MATLAB and SIMULINK, and the charging circuit simulation was completed on PROTEUS platform. Finally, the process of charging lithium batteries was verified via hardware debugging. The results show that the rapid and efficient charging of lithium battery can be achieved from the solar photovoltaic cells by PSO BP neural network system.

关 键 词：最大功率点跟踪 粒子群优化 BP神经网络 脉冲宽度调制 锂电池快速充电 

分 类 号：TM615[电气工程—电力系统及自动化]