生态环境学报 ›› 2025, Vol. 34 ›› Issue (1): 135-144.DOI: 10.16258/j.cnki.1674-5906.2025.01.015

• 研究论文【环境科学】 • 上一篇    下一篇

基于改进CRITIC-云模型法的机场鸟击风险评价

刘振江1(), 王正2, 曹玉洁3, 蔡姝崴3, 陈伟3,*(), 路凯丽3   

  1. 1.海军勤务学院海防工程系，天津 300450
    2.海军航空大学青岛校区航空装备保障指挥系，山东 青岛 266401
    3.海军勤务学院基础部，天津 300450
  • 收稿日期:2024-09-04 出版日期:2025-01-18 发布日期:2025-01-21
  • 通讯作者: * 陈伟。E-mail: 13752220829@163.com
  • 作者简介:刘振江（1975年生），男，教授，硕士，研究方向为机场鸟击防范。E-mail: macrop@163.com
  • 基金资助:
    国家重点研发计划项目(2022YFC3102901)

Risk Assessment of Bird Strike at Airport Based on Improved CRITIC-cloud Model

LIU Zhenjiang1(), WANG Zheng2, CAO Yujie3, CAI Shuwei3, CHEN Wei3,*(), LU Kaili3   

  1. 1. Department of Coastal Defence Engineering, Naval Logistics Academy, Tianjin 300450, P. R. China
    2. Department of Aviation Equipment Support and Command at Qingdao Campus, Naval Aviation University, Qingdao 266401, P. R. China
    3. Department of Basic, Naval Logistics Academy, Tianjin 300450, P. R. China
  • Received:2024-09-04 Online:2025-01-18 Published:2025-01-21

RichHTML

PDF

摘要：

选取我国东部沿海某机场周围8 km范围作为调查区域，根据土地类型进行样线与样点调查法，全年记录到140种鸟类数据。通过对影响鸟击的显性条件和隐性条件分析，确定鸟类集群数量、平均体质量、飞行高度、网捕数量、飞行速度等5个指标为风险评价指标，建立鸟击风险评价指标体系，借助各指标数据特征以区间形式制定等级划分标准。针对鸟情数据冗余与极端值情况，对传统的CRITIC法和云模型法进行改进，得到改进CRITIC-云模型法，并运用到机场鸟击风险评价中。结合鸟类指标数据与传统CRITIC-云模型法和传统CRITIC-云模型L法（传统云模型法）评价结果进行对比分析，发现改进CRITIC-云模型法避免了模型受单个等级区间的影响，能够全面性、综合性评价鸟类的风险等级，评价结果更加准确，与鸟击风险评价方面适配性更高。确定风险评价等级IV级与V级的鸟类为高危鸟种，共26种，其中1种为国家一级重点野生保护动物，5种为国家二级重点野生保护动物。通过对高危鸟种活动环境进行分析，将其归纳为海洋活动鸟类、海洋与陆地活动鸟类、陆地活动鸟类，分别从鸟类性状的生活习性、繁殖特征、生物学特性等3个方面进行阐述，结合机场周边环境制定有效的防范措施。

关键词: 鸟击风险评价, 改进CRITIC法, 云模型, 最大隶属度, 高危鸟种, 防范措施

Abstract:

An 8 km radius surrounding a certain airport in the eastern coastal region of China was selected as the survey area. Using line transects and point count investigation methods based on different land types, 140 bird species were recorded throughout the year. A detailed and comprehensive analysis was conducted on various factors influencing bird strike incidents at airports, and this study deeply studied both the explicit and implicit factors resulting in bird strikes. Through rigorous selection and evaluation, five indices were identified as risk assessment factors: bird cluster quantity, average body weight, flight altitude, net capture quantity, and flight speed. Based on these indices, a bird strike risk assessment index system was established, with index data classified into different levels based on reference materials and aligned with data characteristics. The classification levels of these indices are presented in interval form. Redundancy was observed in the data collected from the field-surveyed birds. The numerical characteristics of the traditional cloud model method were redefined to address the changes in risk values within a single-level interval. The Interval boundary values were applied as expected values for the cloud model. When multiple values existed at the interval boundaries, the expected value was revalidated based on the discrete properties of the cloud model. The strength of the cloud model was fully leveraged to address the fuzziness and uncertainty of the quantitative indices, leading to the development of an improved cloud model method. Additionally, extreme values were observed in the surveyed bird data. When the mean deviation is applied to the data, it can balance the comprehensiveness and extreme values of the data. Thus, the mean deviation was employed instead of the traditional standard deviation in the CRITIC method combined with the entropy weight method to avoid the weight value being affected by a single method. Based on this development, an improved CRITIC method was developed. By integrating the improved CRITIC method with the improved cloud model method, an improved CRITIC-cloud model method was introduced. The improved CRITIC cloud model method was applied to bird strike risk assessment at the airport. Each bird species was assigned a risk evaluation level by evaluating the risk levels of the 140 recorded bird species. A comparison of the specific data for each bird index between the improved method and the traditional CRITIC cloud model L method (traditional cloud model method) revealed that the improved CRITIC cloud model method effectively resolved data redundancy within intervals and minimized the influence of single-level intervals. This provided a more comprehensive and balanced evaluation of bird strike risk levels. In general, the improved CRITIC cloud model method proved to be more accurate and effective, making it well suited for bird strike risk assessments. From the evaluation results, bird species classified as risk levels IV and V were identified as high-risk out of a total of 26 species. These species have a higher probability of bird-strike incidents and could cause significant threats to aviation safety, heightening extra attention. Among these high-risk species, one was classified as a wild animal under Grade I conservation, and five species were classified as Grade II. Further analysis of the habitats and surrounding environments of these high-risk bird species categorized them as marine, marine-terrestrial, and terrestrial birds. This study elaborates on their living habits, reproductive characteristics, and biological traits and proposes effective preventive measures based on the surrounding environment of airports.

Key words: bird strike risk assessment, improved CRITIC method, cloud model, maximum membership degree, high-risk bird species, preventive measures

中图分类号: