华南理工大学学报(自然科学版) ›› 2021, Vol. 49 ›› Issue (9): 56-64.doi: 10.12141/j.issn.1000-565X.200695

所属专题: 2021年化学化工

• 化学化工 • 上一篇    下一篇

基于鸡群算法的危险化学品泄漏扩散事故反演方法

陈国华1,3 胡雪薇1,3 周利兴1,3 门金坤1,3 陈清光2,3   

  1. 1.华南理工大学 机械与汽车工程学院,广东 广州 510640;2.广东省安全生产科学技术研究院,广东 广州 510640;
    3.广东省安全生产科技协同创新中心,广东 广州 510640
  • 收稿日期:2020-11-13 修回日期:2021-01-07 出版日期:2021-09-25 发布日期:2021-09-01
  • 通信作者: 陈国华(1967-),男,教授,博士生导师,主要从事工业安全与风险评价技术、过程装备安全可靠性及风险评价技术研究。 E-mail:mmgchen@scut.edu.cn
  • 作者简介:陈国华(1967-),男,教授,博士生导师,主要从事工业安全与风险评价技术、过程装备安全可靠性及风险评价技术研究。
  • 基金资助:
    广东省科技计划项目(2018B020207010);广州市民生科技攻关项目(201903010102)

Inversion Method Based on Chicken Swarm Optimization for Hazardous Chemicals Leakage and Diffusion Accident

CHEN Guohua1,3 HU Xuewei1,3 ZHOU Lixing1,3 MEN Jinkun1,3 CHEN Qingguang2,3   

  1. 1.School of Mechanical and Automotive Engineering,South China University of Technology,Guangzhou 510640,Guangdong,
    China;2.Guangdong Academy of Safety and Technology,Guangzhou 510640,Guangdong,China;3.Guangdong Provincial 
    Science and Technology Collaborative Innovation Center for Work Safety,Guangzhou 510640,Guangdong,China
  • Received:2020-11-13 Revised:2021-01-07 Online:2021-09-25 Published:2021-09-01
  • Contact: 陈国华(1967-),男,教授,博士生导师,主要从事工业安全与风险评价技术、过程装备安全可靠性及风险评价技术研究。 E-mail:mmgchen@scut.edu.cn
  • About author:陈国华(1967-),男,教授,博士生导师,主要从事工业安全与风险评价技术、过程装备安全可靠性及风险评价技术研究。
  • Supported by:
    Supported by the Science and Technology Planning Project of Guangdong Province (2018B020207010)

摘要: 化工厂危化品泄漏扩散事故时有发生,如何快速精确地确定泄漏源的位置和源强是处置突发危化品泄漏扩散事故的关键。针对传统泄漏源反演方法耗时长、精确度低等缺陷,文中提出了一种基于鸡群算法的泄漏扩散事故反演方法。该方法以气体浓度传感器监测的气体浓度数据与模型计算值的误差为优化目标,将危化品泄漏扩散事故反演问题转化为一个优化目标最小化问题展开研究,利用鸡群算法进行问题求解。采用加权质心法构造高质量的初始解集提升算法精度,反演结果最大误差减小至1m。模拟结果表明,鸡群算法可以在短时间内得出较为准确的泄漏源位置,且鲁棒性满足实际需求,能够实现化工厂危险化学品泄漏源的快速、准确定位,以及泄漏源强的准确预估,为应急辅助决策提供及时有效的支持。

关键词: 危险化学品, 泄漏, 气体扩散模型, 泄漏源反演, 鸡群算法

Abstract: Dangerous chemicals leakage and diffusion accidents occur occasionally in the chemical plant.The key to dealing with such accidents is to confirm the location and rate of leakage sources quickly and accurately.Aiming at the defects of traditional leakage source inversion method,which is time-consuming and low accuracy,this paper proposes a leakage diffusion accident inversion method based on chicken swarm optimization (CSO) algorithm.This method takes the error between the gas concentration data monitored by the gas sensor and the calculated value of the model as the optimization objective.It transforms the inversion problem of hazardous chemical leakage and diffusion accident into an optimization objective minimization problem,and solves this problem with the chicken swarm optimization algorithm.In order to improve the accuracy of the algorithm,a weighted centroid method is used to construct a high quality initial solution set,which achieves an accuracy of 1m.The simulation results show that the chicken swarm optimization algorithm can accurately find the leakage source location in a short time,and its robustness can meet the actual demand; and that the method can realize the rapid and accurate positioning of hazardous chemical leakage sources in the chemical park and the accurate estimation of the strong leakage sources,thus providing timely and effective support for emergency decision-making.

Key words: hazardous chemicals, leakage, gas diffusion model, leakage source inversion, chicken swarm

中图分类号: