摘要
针对空管安全风险预警指标体系构建中,因存在人为认知的局限与风险客观存在之间的矛盾,指标间易出现相关性及冗余性等特征,运用粗糙集在数据挖掘方面的优势,通过属性约简剔除冗余指标,可以实现在不丢失关键预警监控对象的情况下,降低预警建模难度,提高预警决策精度。经过实例运算,将26个监控指标提炼为6个重点观测对象,为管控决策提供了依据。
In the construction of air traffic management safety risk early-warning index system,due to existence of the contradiction between human cognitive limitations and objective risk,it is easy to appear correlation and redundancy among indices.Using rough sets advantages in data mining,through the attribute reduction eliminate redundant index,it can reduce warning modeling difficulties of construction and raise warning decision accuracy in the case of not losing key warning monitoring objects.After example simulation,the 26 monitoring index were refining into 6 key observation objects,which provides the basis for the control of decision-making.
出处
《武汉理工大学学报(信息与管理工程版)》
CAS
2012年第6期776-780,共5页
Journal of Wuhan University of Technology:Information & Management Engineering
基金
国家自然科学基金资助项目(70971104)