GA-BP网络在木材干燥过程建模中的应用

木材干燥是一个复杂的非线性系统,由于木材结构复杂且具有多样性和变异性,很难建立一个理想的符合木材干燥过程的数学模型。利用遗传算法的全局寻优能力优化BP神经网络连接权值系数,分别用BP和GA-BP两种算法建立了木材干燥基准模型。对比结果表明:GA-BP算法建立木材干燥基准模型提高了期望误差精度和收敛速度,避免了BP算法陷入局部极小值,预测平均误差为1.0413%,具有较好的预测精度。

Application on the Identification of Wood Drying process Based on GA-BP

For wood having variety,complexity and variability,wood drying process is a complicated nonlinear system,so it is difficult to get an ideal model for wood drying.The initial weights of BP neural network were evolved by the characteristics of global optimization of Genetic Algorithm,Lumber Moisture Content models are obtained with BP arithmetic and GA-BP arithmetic in this paper.Lumber Moisture Content models were obtained with BP arithmetic and GA-BP arithmetic,Training results showed that meansquare errors and accelerate of convergence were improved with GA-BP lumber moisture Content models,BP arithmetic immersion minim value was avoided,the prediction mean errors were 1.0413% and showed relatively high prediction precision.