一种改进灰狼算法优化LSSVR的混凝土泵车砼活塞剩余寿命预测方法研究

An Improved Gray Wolf Algorithm to Optimize LSSVR for Residual Life Prediction Method of Concrete Pump Truck Concrete Piston

为了解决混凝土泵车砼活塞因无法及时更换导致设备停机的问题,提出一种改进灰狼算法优化最小二乘支持向量回归(LSSVR)的剩余寿命预测方法,该方法使用差分算法(DE)优化原始灰狼算法(GWO),解决了其容易陷入局部最优解的问题,提高了收敛速度,使用优化后的算法优化最小二乘支持向量回归的两个参数,建立剩余寿命预测模型。通过真实的砼活塞寿命监测数据,使用3种评估指标对比LSSVR、GWO-LSSVR、DE-GWO-LSSVR这3个模型的预测效果,并与相关研究的结果进行对比。实验表明,DE-GWO-LSSVR模型拥有最高的预测精度,可以为砼活塞的预测性更换以及机械零件的故障诊断提供指导意义。

In order to solve the problem that the concrete piston of the concrete pump truck could not be replaced in time,an improved gray wolf algorithm was proposed to optimized the remaining life prediction method of the Least Square Support Vector Regression(LSSVR).The new method DE-GWO-LSSVR uses the difference evolution algorithm to optimize the original gray wolf algorithm which solves the problem that it is easy to fall into the local optimal solution,improves the convergence speed,and then uses the optimized algorithm to optimize the two parameters of the least square support vector regression to establish the remaining life prediction model.Based on real concrete piston life monitoring data,three evaluation indicators are used to compare the prediction effects of the three models of LSSVR,GWO-LSSVR,and DE-GWO-LSSVR,and compare them with the results of related studies.Experiments show that the DE-GWO-LSSVR model has the highest prediction accuracy,which can provide guidance for the predictive replacement of concrete pistons and the fault diagnosis of mechanical parts.