多变量预测模型模式识别(variable predictive model based class discriminate,简称VPMCD)利用样本特征值内在的相关性来建立特征学习模型,但是当训练样本较少时会导致模型预测不准确,因此提出了基于递归定量分析(recurrence quantific...多变量预测模型模式识别(variable predictive model based class discriminate,简称VPMCD)利用样本特征值内在的相关性来建立特征学习模型,但是当训练样本较少时会导致模型预测不准确,因此提出了基于递归定量分析(recurrence quantification analysis,简称RQA)和投票法多变量预测模型模式识别(voted variable predictive model based class discriminate,简称V-VPMCD)的故障识别方法。该方法利用了递归定量分析对非线性、非平稳信号分析的鲁棒性和样本质量不高时处理的优势,以VPMCD作为分类方法,并用投票法优化了VPMCD方法,提升了算法的稳定性和识别率。对滚动轴承不同程度、不同类型故障的模式识别实验表明,该优化算法具有较高的识别准确率和稳定性。展开更多
In order to improve the control performance of strip rolling mill, theoretical model of the hydraulic gap control(HGC) system was established. HGC system offline identification scheme was designed for a tandem cold st...In order to improve the control performance of strip rolling mill, theoretical model of the hydraulic gap control(HGC) system was established. HGC system offline identification scheme was designed for a tandem cold strip mill, the system model parameters were identified by ARX model, and the identified model was verified. Taking the offline identified parameters as the initial values, online identification using recursive least square was carried out with model parameters changing. For the purpose of improving system robustness and decreasing the sensitivity due to model errors, the HGC system based on generalized predictive control(GPC) was designed, and simulation experiments for traditional controller and GPC controller were conducted. The results show that both controllers acquire good control effect with model matching. When the model mismatches, for the traditional controller, the overshot will increase to 76.7% and the rising time will increase to 165.7 ms, which cannot be accepted by HGC system; for the GPC controller, the overshot is less than 8.5%, and the rising time is less than 26 ms in any case.展开更多
文摘多变量预测模型模式识别(variable predictive model based class discriminate,简称VPMCD)利用样本特征值内在的相关性来建立特征学习模型,但是当训练样本较少时会导致模型预测不准确,因此提出了基于递归定量分析(recurrence quantification analysis,简称RQA)和投票法多变量预测模型模式识别(voted variable predictive model based class discriminate,简称V-VPMCD)的故障识别方法。该方法利用了递归定量分析对非线性、非平稳信号分析的鲁棒性和样本质量不高时处理的优势,以VPMCD作为分类方法,并用投票法优化了VPMCD方法,提升了算法的稳定性和识别率。对滚动轴承不同程度、不同类型故障的模式识别实验表明,该优化算法具有较高的识别准确率和稳定性。
基金Project(51074051)supported by the National Natural Science Foundation of ChinaProject(20131033)supported by the Ph D Start-up Fund of Natural Science Foundation of Liaoning Province,ChinaProject(N140704001)supported by the Fundamental Research Funds for the Central Universities,China
文摘In order to improve the control performance of strip rolling mill, theoretical model of the hydraulic gap control(HGC) system was established. HGC system offline identification scheme was designed for a tandem cold strip mill, the system model parameters were identified by ARX model, and the identified model was verified. Taking the offline identified parameters as the initial values, online identification using recursive least square was carried out with model parameters changing. For the purpose of improving system robustness and decreasing the sensitivity due to model errors, the HGC system based on generalized predictive control(GPC) was designed, and simulation experiments for traditional controller and GPC controller were conducted. The results show that both controllers acquire good control effect with model matching. When the model mismatches, for the traditional controller, the overshot will increase to 76.7% and the rising time will increase to 165.7 ms, which cannot be accepted by HGC system; for the GPC controller, the overshot is less than 8.5%, and the rising time is less than 26 ms in any case.
