基于GA-BP神经网络的轻骨料混凝土弹性模量预测

Prediction of Elastic Modulus of Lightweight Aggregate Concrete Based on GA-BP Neural Network

摘要: 轻骨料混凝土弹性模量的预测能够优化材料配比、降低工程成本、提升结构安全性和推动绿色建筑发展，具有重要的意义。应用遗传算法(GA)优化BP神经网络的网络结构及网络权重和阈值，使其拥有更高的精度。通过GA-BP神经网络模型预测轻骨料混凝土的弹性模量，并对模型进行评估和应用。结果表明，GA-BP神经网络模型训练完成后的训练集、验证集、测试集和完整数据集的目标间相关度均在0.94以上；GA-BP神经网络模型预测结果所得的RMSE值和MAE值均低于BP神经网络；GA-BP神经网络模型应用所得的预测值与真实值间平均误差为0.759。GA-BP神经网络能够较为准确地预测轻骨料混凝土的弹性模量。

Abstract: The prediction of the elastic modulus of lightweight aggregate concrete holds significant importance for optimizing material mix ratios, reducing engineering costs, enhancing structural safety, and promoting green building development. This study applies the Genetic Algorithm (GA) to optimize the network structure, weights, and thresholds of the BP neural network, thereby improving its prediction accuracy. The GA-BP neural network model is employed to predict the elastic modulus of lightweight aggregate concrete, followed by model evaluation and application. Results demonstrate that the correlation coefficients between targets in the training set, validation set, test set, and complete dataset all exceed 0.94 after GA-BP model training. The Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) values obtained from the GA-BP neural network predictions are lower than those of the conventional BP neural network. The average error between predicted values and actual measurements in model applications reaches 0.759. These findings indicate that the GA-BP neural network can achieve relatively accurate predictions for the elastic modulus of lightweight aggregate concrete.