基于神经网络算法的钒基镍氢电池负极材料性能优化  

Performance Optimization of Negative Electrode Material for Vanadium-based Nickel Metal Hydride Battery Based on Neural Network Algorithm

作　　者：王红[1] 王群群 Wang Hong;Wang Qunqun(School of Mechanical and Electrical Engineering,Guangzhou Institute of Technology,Guangzhou 510075,Guangdong,China;School of Materials Science and Engineering,Chongqing University of Science and Technology,Chongqing 400054,China)

机构地区：[1]广州工程技术职业学院机电工程学院,广东广州510075 [2]重庆理工大学材料科学与工程学院,重庆400054

出　　处：《钢铁钒钛》2020年第6期60-65,共6页Iron Steel Vanadium Titanium

基　　金：广东省教育厅2017年度科研平台和科研项目之特色创新项目(自然科学)(2017GKTSCX052)。

摘　　要：采用6×36×12×1四层拓扑结构神经网络模型,以钛含量、镍含量、铝含量、铬含量、保温温度和保温时间为输入层参数,以充放电循环稳定性为输出层参数,构建了钒基镍氢电池负极材料性能优化神经网络模型。结果表明,模型具有较强预测能力和较高预测精度,平均相对训练误差4.8%、平均相对预测误差4.9%。与现有V3Ti Ni0.56材料相比,神经网络模型优化获得的V3Ti Ni0.56Al0.3Cr0.4材料在充放电循环30次后的放电容量衰减率从61%减小到26%。The neural network model with 6×36×12×1 four-layer topological structure was used to optimize the performances of the negative electrode material of vanadium based hydrogen storage battery.The input layer parameters were titanium content,nickel content,aluminum content,chromium content,holding temperature and holding time.The output layer parameters were charge and discharge cycle stability.The results show that the model has strong prediction ability and high prediction accuracy,with an average relative training error of 4.8%and an average relative prediction error of 4.9%.Compared with the existing V3Ti Ni0.56material,the capacity decay rate of V3Ti Ni0.56Al0.3Cr0.4material optimized by neural network model after 30 charge-discharge cycles is reduced from 61%to 26%.

关 键 词：钒基镍氢电池 负极材料 V3TiNi0.56Al0.3Cr0.4 神经网络 充放电循环稳定性 

分 类 号：TF841.3[冶金工程—有色金属冶金] TM911[电气工程—电力电子与电力传动]