基于BP神经网络模型的硅灰–电石渣协同固化黏土强度仿真与预测

Simulation and Prediction of Strength of Silica Fume-Carbide Slag Synergistically Stabilizing Clay Based on BP Neural Network Model

摘要: 硅灰–电石渣协同加固黏土是一种新型的工业固废再利用在土体加固的方式，为了减少室内试验工作量以及预测固化土的无侧限抗压强度，本研究通过构建BP神经网络模型对其进行预测。通过测试，模型的最佳均方误差为0.116，训练组、验证组和测试组的相关拟合系数R都在0.9以上，附加预测值和真实值较为接近，由此说明神经网络模型可靠，对预测硅灰–电石渣固化土的无侧限抗压强度较为精准，是一种可以降低成本减少试验工作量的预测方法。

Abstract: The synergistic reinforcement of clay using silica fume and carbide slag is a novel approach for the reuse of industrial waste in soil stabilization. To reduce the laboratory testing workload and predict the unconfined compressive strength of the stabilized soil, this study constructs a BP neural network model for prediction. The testing results show that the model’s optimal mean square error is 0.116, with the correlation coefficients (R) for the training, validation, and testing sets all exceeding 0.9. The predicted values are close to the actual values, demonstrating that the neural network model is reliable and provides accurate predictions for the unconfined compressive strength of silica fume-carbide slag stabilized soil. This approach offers a cost-effective and efficient method for reducing experimental workload.