基于PSO-SVM算法的梯级泵站管道振动响应预测

发布时间：2018-06-04 04:14

本文选题：泵 + 振动　；参考：《农业工程学报》2017年11期

【摘要】：泵站管道振动响应信号实测比较困难,为实现利用较少机组数据预测管道振动状况,提出基于粒子群(particle swarm optimization,PSO)的支持向量机(support vector machine,SVM)预测方法。利用粒子群全局跟踪搜索算法优化SVM核函数和惩罚因子,弱化SVM参数优化不足导致预测精度低的问题。以景电梯级二期3泵站2号管道为研究对象,基于机组和管道的振动实测数据,首先利用频谱分析和数理统计方法确定管道振动的振源贡献率,并计算机组和管道振动相关系数,确定机组和管道之间的强耦合关系。然后建立泵站管道振动的PSO-SVM预测模型,选取机组不同时段振动实测数据作为输入因子,相应时段管道振动数据作为输出因子进行训练和振动预测,并将管道振动预测结果与BP神经网络预测结果进行对比。与BP网络神经预测结果相比,该方法预测结果与实测值吻合度高,其平均相对误差最大为6.8%,根均方误差最大为0.261,预测精度更高。能够有效实现管道的振动响应预测,从而达到管道实时在线安全运行监测的目的。
[Abstract]:It is difficult to measure the vibration response signal of pumping station pipeline. In order to use less data to predict pipeline vibration, a prediction method of support vector machine (SVM) based on particle swarm optimization (PSO) is proposed. The particle swarm optimization algorithm is used to optimize the SVM kernel function and penalty factor, and to weaken the problem of low prediction precision caused by the insufficient optimization of SVM parameters. Based on the measured vibration data of the unit and pipeline, the contribution rate of pipeline vibration source is determined by spectrum analysis and mathematical statistics method. The strong coupling relationship between the unit and the pipeline is determined by the correlation coefficient between the computer group and the pipeline vibration. Then the PSO-SVM prediction model of pipeline vibration in pumping station is established. The measured vibration data of unit in different periods are selected as input factors, and pipeline vibration data in corresponding period are used as output factors for training and vibration prediction. The prediction results of pipeline vibration are compared with those of BP neural network. Compared with the BP neural network prediction results, the prediction results of this method are in good agreement with the measured values. The average relative error is 6.8. the root mean square error is 0.261, and the prediction accuracy is higher. It can effectively realize the prediction of pipeline vibration response and achieve the purpose of real-time on-line safe operation monitoring.
【作者单位】：华北水利水电大学水利学院;
【基金】：国家自然科学基金(51679091) 华北水利水电大学研究生教育创新计划基金(YK2015-02)
【分类号】：TP18;TV675

【参考文献】