论文部分内容阅读
以基于红外热成像机理测量真空绝热板(VIP)的真空度为研究对象。鉴于目前此方案测量真空度存在精度低的问题,本文提出一种基于BP神经网络的真空度测量精度改进方法。针对BP神经网络存在易陷入局部最优、收敛速度慢等缺陷,本文巧妙利用思维进化算法优化BP神经网络的初始权值和阈值,从而弥补以上缺陷。最终实验结果表明,利用思维进化算法优化BP神经网络创建的数学模型大大提高了真空绝热板真空度的测量精度,其实际测量精度优于2%。因此,本文所提方法具有广泛推广的应用价值。
The vacuum degree of vacuum insulation panel (VIP) was measured based on infrared thermal imaging mechanism. In view of the problem of low accuracy of vacuum measurement in this scheme, an improved method of vacuum measurement accuracy based on BP neural network is proposed in this paper. Due to the defects of BP neural network, such as easy to fall into local optimum and slow convergence speed, this paper cleverly uses the thought evolutionary algorithm to optimize the initial weights and thresholds of BP neural network, so as to make up for these shortcomings. The final experimental results show that the optimization of the neural network based on thought evolutionary algorithm can greatly improve the measurement accuracy of the vacuum insulation panel, and the actual measurement accuracy is better than 2%. Therefore, the method proposed in this article has wide application value.