基于IWOA-LMBP的水稻插秧机可靠性预测  

Reliability Prediction of Rice Transplanter Based on IWOA-LMBP

作　　者：文昌俊[1,2] 陈凡 康锁 陈洋洋 WEN Changjun;CHEN Fan;KANG Suo;CHEN Yangyang(School of Mechanical Engineering,Hubei Univ.of Tech.,Wuhan 430068,China;Hubei Key Laboratory of Modern Manufacturing Quality Engineering,Wuhan 430068,China)

机构地区：[1]湖北工业大学机械工程学院,湖北武汉430068 [2]湖北省现代制造质量工程重点实验室,湖北武汉430068

出　　处：《湖北工业大学学报》2025年第2期1-10,共10页Journal of Hubei University of Technology

基　　金：国家自然科学基金(51875180)。

摘　　要：针对水稻插秧机可靠性评价存在明显滞后的问题,提出一种创新的解决方案:构建改进鲸鱼算法-列文伯格·马夸德优化的BP神经网络可靠性预测模型。其设计思路如下:首先,引入Chebyshev混沌策略,以增强初始种群的多样性;其次,采用“双阶梯”和“双山谷”非线性自适应因子,动态平衡算法的全局搜索与局部勘探能力;最后,结合趋优透镜反向学习策略,以更新个体位置,进一步提升个体质量,有效帮助算法跳出局部最优。通过6个基准测试函数的寻优对比分析和Wilcoxon秩和统计检验可知,IWOA具有更好的寻优性能。随后,利用现场跟踪获取的水稻插秧机故障数据,建立IWOA-LMBP模型。为了全面评估该模型的性能,选取MAE、RMSE、R 2作为网络模型的评价指标,并将其与其他5种模型进行对比。结果表明:采用IWOA-LMBP模型进行预测时,效果更好。该IWOA-LMBP模型为水稻插秧机可靠性精准预测提供了参考依据。Aiming at the problem that the reliability evaluation of Rice transplanter lags behind,an improved Whale Optimization Algorithm(IWOA)-Levenberg Marquard(LM)BP neural network reliability prediction model was established.First,the initial population diversity is enhanced through the Chebyshev chaotic strategy;Second,the"double ladder"and"double valley"nonlinear adaptive factors are used to achieve the global search and local exploration capabilities of the Dynamic equilibrium algorithm;Finally,combining the optimization lens reverse learning strategy to update individual positions,improve individual quality,and help the algorithm jump out of local optima.Through comparative analysis of six benchmark test functions and Wilcoxon rank sum statistical test,it can be concluded that IWOA has better optimization performance.Then,the IWOA-LMBP model is established based on the fault data of the Rice transplanter tracked on site.Selecting MAE and RMSE and R 2 as evaluation indicators for the network model,and comparing them with the other five models,the results show that the IWOA-LMBP model performs better in prediction.The model can provide a reference for the accurate prediction of Rice transplanter.

关 键 词：水稻插秧机 改进鲸鱼算法 趋优透镜反向学习 LMBP神经网络 可靠性预测 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]