基于GA-SVM的车轮磨耗数据预测  

Prediction of Wheel Wear Data Based on GA-SVM

作　　者：李喆 潘鸿飞 牛泓博 张旭[2,3] 

机构地区：[1]中车大连机车车辆有限公司,辽宁 大连 [2]大连交通大学机车车辆工程学院,辽宁 大连 [3]大连交通大学高速列车服役安全关键技术铁路行业重点实验室,辽宁 大连

出　　处：《数据挖掘》2024年第1期20-25,共6页Hans Journal of Data Mining

摘　　要：在列车车辆运行过程中,列车车轮承担着重要的角色,它们负责支撑整个列车的负重,保障列车安全,于是预测车轮磨损的研究变得极为重要。随着大数据分析技术的不断发展,各种智能算法逐渐被引入车轮磨损的预测中,以提高预测的准确性。在此背景下,本研究采用遗传算法和支持向量机模型对车轮磨损的回归预测。预测结果的RMSE的值仅为0.059,说明该模型具有优秀的预测效果。During the operation of train vehicles, train wheels play an important role in supporting the weight of the entire train and ensuring train safety. Therefore, research on predicting wheel wear has be-come extremely important. With the continuous development of big data analysis technology, various intelligent algorithms are gradually being introduced into the prediction of wheel wear to im-prove the accuracy of prediction. In this context, this study uses genetic algorithms and support vector machine models for regression prediction of wheel wear. The RMSE value of the predicted result is only 0.059, indicating that the model has excellent predictive performance.

关 键 词：车轮磨耗 支持向量机 遗传算法 大数据技术 

分 类 号：G63[文化科学—教育学]