SVM的起重机金属结构安全评估研究

Metal Structure Safety Assessmentof Crane Based on SVM

通过整理现有桥式起重机规范文件及结合工程经验,提出以量化的方式处理桥式起重机检测数据样本,及以PCA(Principal Component Analysis)降维算法优化样本。利用SVM(支持向量机)算法建立桥式起重机金属结构安全评估系统,对桥式起重机金属结构进行智能安全评估。通过Matlab编程工具,开发了桥式起重机金属结构安全评估程序,结合工程检测样本进行仿真实验分析。实验结果表明,针对桥式起重机金属结构安全评估这类小样本、非线性问题,本评估方法具有较高的评估可靠性,可为桥式起重机金属结构安全评估提供技术支持。

By arranging code documents on bridge crane and combining with engineering experience, quantitative sample method and PCA （ Prineipa! Component Analysis ）dimension reduction algorithm dealing with test-samples of bridge crane has been put forward. Through SVM （support vector machine ）, safety assessment system of bridge crane metal structure has been established, and intelligence security assessment has been done. Meanwhile, safety assessment process in metal structure of bridge crane has been implemented through matlab. Then case study has been carried on according to detection samples. Results show that referring to this small sample and non-linear issue, this assessment method is of high estimate reliability and can provide technical support to conduct scfety assessment on metal structure of bridge crane.