基于人工蜂鸟算法与三参数威布尔分布的机械密封可靠性评估

RELIABILITY ESTIMATION OF MECHANICAL SEALS BASED ON ARTIFICIAL HUMMINGBIRD ALGORITHM AND THREE⁃PARAMETER WEIBULL DISTRIBUTION

为了对机械密封件进行快速准确的可靠性评估,采用三参数威布尔(Weibull)分布对机械密封故障数据进行建模,将机械密封可靠性评估问题转化为约束优化问题,提出基于人工蜂鸟算法(Artificial Hummingbird Algorithm,AHA)的可靠性评估方法,并通过数值模拟和实例分析比较了极大似然估计(Maximum Likelihood Estimation,MLE)法、改进哈里斯鹰优化(Improved Harris Hawks Optimization,IHHO)算法和AHA在求解Weibull分布参数估计问题中的计算精度及运行效率。结果表明,基于AHA进行参数估计具有收敛速度快和计算精度高的优点,较MLE和IHHO算法寻优时间缩短且精度提高,从而为机械密封的可靠性评估提供了一种新的高效方法。

In order to quickly and accurately evaluate the reliability of mechanical seals,the three⁃parameter Weibull distribution was used to model the mechanical seal failure data.The reliability evaluation problem of mechanical seal was transformed into a constraint optimization problem,and a reliability evaluation method based on artificial hummingbird algorithm(AHA)was proposed.Through numerical simulation and case analysis,maximum likelihood estimation(MLE),improved Harris Hawks optimization(IHHO)algorithm and AHA were used to compare calculation accuracy and operational efficiency in solving the Weibull distribution parameter estimation problems.The results show that the parameter estimation using AHA has advantages of faster convergence speed and higher accuracy,and it can shorten the optimization time and improve the accuracy compared with MLE algorithm and IHHO algorithm,which provides a new and efficient method for Weibull parameter estimation of mechanical seals.