| 基于支持向量机的化工工艺安全评价模型构建及优化研究* |
| |
| 引用本文： | 刘晋,程彦斌,齐东川,蒋晋,王文和.基于支持向量机的化工工艺安全评价模型构建及优化研究*[J].中国安全生产科学技术,2022,18(12):154-161. |
| |
| 作者姓名： | 刘晋 程彦斌 齐东川 蒋晋 王文和 |
| |
| 作者单位： | （1.重庆科技学院 安全工程学院，重庆 401331；2.重庆市安全生产科学研究院，重庆 401121；3.西南兵工重庆环境保护研究所有限公司，重庆 400042） |
| |
| 基金项目： | * 基金项目: 教育部人文社会科学研究青年基金项目(20YJCZH096);国家应急管理部安全生产重大事故防治关键技术科技项目（chongqing-0004-2018AQ）;重庆市教育委员会科学技术研究项目(KJZD-K202101501);油气生产安全与风险控制重庆市重点实验室 开放基金项目（cqsrc202114）;重庆市璧山区科学技术局科研项目（BSKJ20210027） |
| |
| 摘 要： | 为解决现阶段基于风险分级的安全评价方法仍存在着高维数据处理不当、评价智能化程度不高等问题，创建支持向量机的安全评价模型，利用核函数解决安全评价因子分类问题，粒子群算法（PSO）寻找最适合模型的正则项C，进一步提升安全评价模型的正确率，形成适用高维数据的化工工艺安全评价方法。研究结果表明:该模型与经典支持向量机模型和BP神经网络评价模型相比具有更高的正确率，研究结果对借用机器学习来创新安全评价理论及工程应用具有现实意义及理论价值。
|
| 关 键 词： | 化工工艺 安全评价 风险分级 支持向量机 |
Research on construction and optimization of chemical process safety evaluation model based on support vector machine |
| |
| LIU Jin,CHENG Yanbin,QI Dongchuan,JIANG Jin,WANG Wenhe.Research on construction and optimization of chemical process safety evaluation model based on support vector machine[J].Journal of Safety Science and Technology,2022,18(12):154-161. |
| |
| Authors: | LIU Jin CHENG Yanbin QI Dongchuan JIANG Jin WANG Wenhe |
| |
| Affiliation: | (1.College of Safety Engineering,Chongqing University of Science & Technology,Chongqing 401331,China;2.Chongqing Academy of Safety Science and Technology,Chongqing 401121,China;3.Southwest Armaments Chongqing Environmental Protection Research Institute Co.,Ltd.,Chongqing 400042,China) |
| |
| Abstract: | In order to solve the problems of improper processing of high-dimensional data and low intelligence degree of evaluation in the current safety evaluation methods based on risk classification,a safety evaluation model of support vector machine was created.The kernel function was used to solve the classification problem of safety evaluation factors,and the particle swarm optimization algorithm (PSO) was used to find the most suitable regular term C of the model,so as to further improve the accuracy of the safety evaluation model,and an evaluation method of chemical process safety suitable for high-dimensional data was formed.The results showed that the model had 92% accuracy compared with the classical support vector machine model and BP neural network evaluation model.It has practical significance and theoretical value for using machine learning to innovate the safety evaluation theory and engineering application. |
| |
| Keywords: | chemical process safety evaluation risk classification support vector machine |
|
| 点击此处可从《中国安全生产科学技术》浏览原始摘要信息 |
|
点击此处可从《中国安全生产科学技术》下载全文 |
