基于径向基网络的炭黑填充胶生热率模型重构

MODEL RECONSTRUCTION OF HEAT GENERATION RATIO OF CARBON BLACK-FILLED RUBBER BASED ON RBF NN

在分析炭黑填充胶生热率随温度、频率变化趋势基础上,构建了生热率的径向基网络模型(RBF网络),并对RBF网络进行模拟,将RBF网络与反向传播网络(BP网络)进行比较,分析误差。模拟结果表明RBF网络精度最高,探讨了将人工神经网络理论引入到炭黑/橡胶复合材料热物性参数研究领域的可行性。

An radial basis function neural network（RBF NN） model is established based on the analysis of the relation between temperature, frequency and heat generation rate of carbon black-filled rubber. The RBF NN model is simulated and is compared with the BP neural network model, and their relative errors are analyzed. The simulation results indicate the RBF NN＇ precision is greater than the BP NN＇s, and artificial neural network theory can be used in the research field of carbon black-filled rubber＇s thermophysical properties.