论文部分内容阅读
鉴于大坝渗流监测受众多因素影响,首先利用主成分分析法对相关性较大的因子进行处理,然后利用最小二乘支持向量机进行建模,最后依靠遗传算法对其参数进行选优,建立了基于改进的最小二乘支持向量机的闸首渗流监控模型,并通过实例应用做了对比分析。结果表明,改进的最小二乘支持向量机模型可有效降低输入因子的维数,减小因子之间相关性,降低模型的训练时间,拟合精度均优于其他模型,更适合于渗流监测数据的建模。
In view of many factors affecting the seepage monitoring of the dam, the principal component analysis method is used to deal with the more relevant factors, and then the least square support vector machine is used to model. Finally, genetic algorithm is used to optimize its parameters to establish The sluice seepage monitoring model based on improved least square support vector machine was compared and analyzed by examples. The results show that the improved least square support vector machine model can effectively reduce the dimensionality of input factors, reduce the correlation between factors and reduce the training time of the model, the fitting accuracy is better than other models, and is more suitable for seepage monitoring data Modeling.