基于RS-SCA-PPR的充填管道失效风险预测精度研究

为提高充填管道失效风险的预测精度,建立了基于粗糙集(RS)和正余弦(SCA)算法优化投影寻踪回归(PPR)的充填管道失效风险预测模型。以某矿山为例,选取10项影响因素构建充填管道失效风险预测指标体系,通过RS属性约简原理提取5项核心因素,再利用PPR对充填管道失效风险进行预测,并采用SCA对模型参数进行优化。结果表明:RS可有效消除冗余信息,简化运算过程,SCA-PPR预测精度高、模型性能好,拓宽了矿山充填管道失效风险预测研究的思路。

Precision of Failure Risk Prediction of Filling Pipeline Based on RS-SCA-PPR

In order to improve the precision of failure risk prediction of filling pipeline, a filling pipeline failure risk prediction model based on rough set (RS) and sine cosine algorithm(SCA) was established to optimize projection pursuit regression (PPR). Taking a specific mine as an example, ten influencing factors were selected to construct the failure risk prediction index system of filling pipeline. Five core factors were extracted by RS attribute reduction principle, then PPR was used to forecast the filling pipeline failure risk, and SCA was used to optimize the model parameters. The results show that the RS can effectively eliminate redundant information to simplify the operation process. SCA-PPR model has a higher prediction accuracy and better model performance, which broadens the thinking of mine filling pipeline failure risk prediction research.