基于指数平滑和改进增量SVR的齿轮寿命预测研究

发布时间：2018-12-11 01:37

【摘要】：针对齿轮剩余使用寿命预测这一问题进行了研究,提出了一套基于二次指数平滑和改进增量支持向量回归(Support vector machine for regression,SVR)的齿轮剩余寿命预测新方案。该方案利用主成分分析法(Principal component analysis,PCA)筛选组成融合指标集,再进行二次指数平滑处理(Double exponential smoothing,DES)作为输入,构建基于增量拉丁超立方采样(Latin hyper-cube sampling,LHD)和Direct搜索算法的改进增量SVR预测模型。与基于原始特征值的标准SVR模型相比,新模型克服了信号随机性和突变性的干扰,并获取了最新状态信息的内在演变趋势,寻优效果和预测性能得到明显改善。用齿轮全寿命周期试验数据对新模型进行验证分析,结果表明,改进增量SVR预测模型获得了更可靠、更稳定的预测结果,具有一定工程实用价值。
[Abstract]:Based on quadratic exponential smoothing and improved incremental support vector regression (Support vector machine for regression,SVR), a new scheme of gear residual life prediction is proposed. In this scheme, principal component analysis (Principal component analysis,PCA) is used to select the fusion index set, and then quadratic exponential smoothing (Double exponential smoothing,DES) is used as input to construct incremental Latin hypercube sampling (Latin hyper-cube sampling,). LHD) and improved incremental SVR prediction model of Direct search algorithm. Compared with the standard SVR model based on the original eigenvalue, the new model overcomes the interference of signal randomness and mutation, and obtains the internal evolution trend of the latest state information. The optimization effect and prediction performance are obviously improved. The new model is validated and analyzed by using the gear life cycle test data. The results show that the improved incremental SVR prediction model has more reliable and stable prediction results and has certain engineering practical value.
【作者单位】：重庆大学机械传动国家重点实验室;
【基金】：国家自然科学基金重点项目(51035008) 机械传动国家重点实验室自主项目(SKLMT-ZZKT-2012MS 02) 重庆市自然科学基金(CSTC,2009BB3365)
【分类号】：TH132.41

【相似文献】