基于改进的最小二乘支持向量机的闸首渗流监控模型研究

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。

Monitoring Model of Ship Lock Head Seepage Based on Improved Least Square Support Vector Machine

The seepage monitoring of dam is influenced by many factors. Firstly, the dimension of the lager correlation factor is reduced by using principal component analysis. And then the model is established with the least squares support vector machine and genetic algorithm is applied to select parameters. Finally, a monitoring model of ship lock head seepage is established. The example results show that the improved least squares support vector machine model can reduce the dimension of input factors and the correlation among the factors and training time effectively; the fitting accuracy is better than other models, which is more suitable for modeling seepage monitoring data.