首页|基于AI的系统适航符合性验证方法研究

基于AI的系统适航符合性验证方法研究

扫码查看
随着人工智能(AI)技术的快速发展,它在民用航空领域的应用日益广泛,同时带来了前所未有的安全风险.因此,研究有效的适航符合性验证方法变得尤为重要.本文首先研究了AI智能系统的定义和等级分类.其次,通过对当前研究成果的梳理,研究了基于创建可信赖AI框架来增强人工智能的可信度,确保AI智能系统适航安全性的方法.最后,研究了构成可信AI框架的各个构建块的实施流程及其技术细节,为民用飞机AI智能系统的设计和认证提供参考依据.
Research on Airworthiness Compliance Verification Methods for AI-based Systems
With the rapid development of artificial intelligence(AI)technology,its application in civil aviation is becoming more and more widespread,meanwhile,it brings unprecedented safety risks.Therefore,it is particularly important to study effective methods for verifying airworthiness compliance.Firstly,the definition and classification of AI intelligent systems are studied.Secondly,through the sorting out of current research results,methods to enhance the credibility of artificial intelligence by creating a trustworthy AI framework to ensure the airworthiness safety of AI intelligent systems are studied.Finally,the implementation process and technical details of each building block that constitutes a trustworthy AI framework are studied,so as to provide reference for the design and certification of AI intelligent systems of civil aircraft.

AI intelligent systemairworthiness compliancetrustworthinessAI assurancelearning assurance

李华、宋海龙、黎明

展开 >

中电科航空电子有限公司,四川 成都 611731

中电科航空电子技术发展研究中心,四川 成都 610036

AI智能系统 适航符合性 可信度 人工智能保证 学习保证

2024

民航学报

民航学报

ISSN:
年,卷(期):2024.8(1)
  • 3