种族鱼群优化支持向量机序列理论监测尾矿坝

为监测预警尾矿坝的变形位移,提出基于结构风险最小化理论的支持向量机进行学习预测。通过采集有效数据,对时间序列数据进行归一化序列处理,然后采取种族鱼群选择向量机参数,对处理后的数据进行支持向量机回归预测。将该理论应用到某尾矿坝监测系统,得到了较为准确的预测结果,表明该理论充分利用了数据的统计特性,精度和泛化能力都得到了明显提高,可作为尾矿坝监测系统的有效指导。

Monitoring Tailings Dam Based on Support Vector Machine Sequence Theory Optimized by Racial Fish

To monitor and early warn tailings dam deformation , it is put forward the theory of support vector machine (SVM ) based on structural risk minimization to study forecasts .Through effective data ,first of all ,time sequence data is processed ,then the racial fish is adopted to choose vector machine parameters and finally the support vector machine (SVM ) is applied to regress and predict the processed data .This theory is applied in the monitoring system of one tailings dam ,the accurate prediction results are obtained ,indicating that the theory makes full use of the statistical properties of the data ,the precision and generalization ability has obviously been improved ,effectively directive to tailings dam monitoring system .