摘要
基于误差反向传播算法,建立发光二极管(LED)散热体冲压成形尺寸和形状误差数学模型.阐述MATLAB神经网络工具箱求解误差数学模型的具体过程.针对某具体LED散热体产品,基于其相关的仿真数据与设计的MATLAB程序,求解出该LED散热体尺寸和形状误差数学模型.验证结果表明:实验的壁厚分布值与文中数学模型预测的壁厚分布值走势整体吻合,局部偏差不大;所建立的LED散热体冲压成形尺寸、形状误差数学模型具有一定的实用性.
Based on error back propagation algorithm,the mathematical model of stamping size and shape error for light emitting diodes(LED)heat dissipation body was established.Explain the specific process to solve the error mathematics model using MATLAB neural network toolbox.Based on a product of LED heat dissipation body,used the related simulation data and the designed MATLAB program to solve the mathematical model of stamping size and shape error for LED heat dissipation body.Verification results showed that the overall value was matched and the local deviation was little for the comparison between the experimental value of the wall thickness distribution and the mathematics model predicted value of that in this paper,the established mathematical model of stamping dimension-error and shape-error for LED heat dissipation body has fine practicality.
出处
《华侨大学学报(自然科学版)》
CAS
北大核心
2015年第4期371-376,共6页
Journal of Huaqiao University(Natural Science)
基金
福建省科技计划项目(2014I01010233)
关键词
误差反向传播
发光二极管
散热体
误差数学模型
壁厚
error back propagation
light emitting diodes
heat dissipation body
error mathematical model
wall thick-ness