镀有石墨烯薄膜的散热器散热性能及其模型的研究

Study on Heat Dissipation Performance and Model of Heak Sink with Graphene Films

为了探究石墨烯薄膜对散热器散热性能的影响,研究了石墨烯薄膜散热器的深度学习模型,采用BP神经网络算法求解石墨烯薄膜散热器散热性能。建立散热器仿真模型,研究发现影响石墨烯薄膜散热器性能的主要因素为功率器件输入功率和功率器件与散热器接触面的面积比,其中随着输入功率的增大和面积比的减小,散热性能变化更加明显。使用输入功率和面积比作为BP神经网络输入特征,功率器件表面温度作为输出特征,优化BP神经网络模型。发现当神经元个数为14个、隐藏层数量为4层时,该网络性能达到最优。代入预测集,最终预测精度可以达到99.8%以上,表明该BP神经网络算法可为选择合适的散热器提供很好的理论依据。并且通过实验发现,镀有石墨烯薄膜的散热器比未镀石墨烯薄膜的散热器的散热性能提升了4.12%。

To explore the effect of graphene film on performance of heat sink,the deep learning model of the graphene film heat sink was studied,and BP neural network algorithm was used to solve the heat dissipation performance of graphene film heat sink.The heat sink simulation model was established.And it is found that the main factors affecting the performance of the graphene film heat sink are the input power of the power device and the area ratio of contact surface between the power device and heat sink.As the input power increases and the area ratio decreases,the heat dissipation performance changes more obviously.With the input power and area ratio as the input feature and the surface temperature of the power device as the output feature,the BP neural network model was optimized.It is found that the performance of the network is optimal when the number of neurons is 14 and the number of hidden layer is 4.When the prediction set is substituted,the final prediction accuracy can reach more than 99.8%,indicating that the network algorithm can provide a good theoretical basis for selecting a suitable graphene film heat sink.And through experiment it is found that the heat dissipation performance of the heat sink with graphene film is 4.12%higher than that of the heat sink without graphene film.