提出了深基坑变形预测的进化支持向量机方法。利用遗传算法来搜索支持向量机与核函数的参数,避免了人为选择参数的盲目性,同时提高了支持向量机的推广预测能力。利用优化后的模型对基坑实例进行了变形预测,并将预测结果与监测结果进行了对比。研究结果表明,该模型与神经网络模型相比,具有预测精度高、泛化能力强等优点,对基坑安全监控具有实用价值。 The evolution support vector machine method for deep excavation deformation prediction is proposed. The genetic algorithm is used to search the parameters of support vector machines and kernel functions, which avoids the blindness of manual selection of parameters and improves the ability of generalized prediction of support vector machines. The optimized model was used to predict the deformation of the foundation pit, and the prediction results were compared with the monitoring results. The results show that compared with the neural network model, the model has the advantages of high prediction accuracy and generalization ability, and has practical value for the safety monitoring of foundation pit.