人员培训对无人机维修差错影响的动态模式与培训收益风险预估研究

Currently, the unmanned aerial vehicle systems (UAS) of China can not satisfy the advised airworthiness regulation requirements, which is not conducive to their civil use and market development in the future. Amount of accident investigations show a trend that “maintenance errors” resulted by “training flaws” contribute to the chief UAS accident causal factors. However, traditional and physical system failure based risk analysis techniques do not view the UAS accidents from the angle of a complex system consisted by “organization-human-physical system”. It cannot distinguish the cause-consequence relationships and the time series of risk factors and also not identify the nonlinear dynamics characteristics of risk coupling effects. These deficiencies make the training investments of maintaining organization pointless and less logical. Based on the system theories, this research studies the risk dynamic evolution process of the training flaws taking effects on the maintenance errors, investigates the typical UAS accident cases in details, identifies the risk factors feedback loops consisted by the organization training activities, maintainer training benefits，maintenance errors and catastrophic accident rates. The method to establish the system dynamics model and promote the qualitative time-domain simulation is proposed and the typical risk dynamics modes of “training flaws-maintenance errors” are concluded. Moreover, an application software called “UAS maintenance training benefit risk estimation program” is developed. It is a utility tool for helping the Chinese UAS maintaining organizations estimate their training benefits in the medium and long term and optimize their training strategies. It will improve the airworthiness of Chinese UAS and protect public safety.

目前我国无人机系统安全性不能满足适航规章建议要求，严重制约其民用化和市场化发展。事故调查表明，“培训缺陷”所导致的“维护差错”是无人机事故的首要致因，传统静态的、基于物理失效的风险分析技术未基于“组织—人员—物理系统”复杂系统视野，不能有效区分风险因素的主次性、时序性，不能识别因素间非线性、动态耦合特征，维护组织的培训投入缺乏针对性且逻辑性不强。本项目拟以“系统理论”视角重新认识培训缺陷导致无人机维护差错的风险动态演化过程，深入分析国内外典型无人机系统的事故案例，辨识由组织培训活动、维护人员培训收益、维护差错及无人机事故率等构成的风险因素反馈回路，建立系统动力学模型开展定性时域仿真，揭示培训缺陷对无人机维护差错影响的典型风险动态模式，开发“无人机系统维护培训收益风险预估程序”软件，为我国无人机维护组织预测培训的中长期收益、优化培训策略提供实用工具，改善我国无人机系统的适航性，保障公众安全。

本项目围绕目前我国无人机系统安全性不能满足适航规章建议要求，严重制约其民用化和市场化发展的现状，通过调查和分析发现维护人因是导致无人机事故的主要致因之一，而传统的人因可靠性研究方法主要采用静态、线性化的安全性模型和视野，忽略因素之间的层次化关系、尤其是人员绩效因子（PsF）之间的反馈式、动态式的作用关系，未能从“组织—人员—物理系统”复杂系统视野全局化应对维护差错和人因风险，对于无人机系统的维护差错抑制效果有限。本项目深入研究了大量的国内外无人机事故、分析了维护人员人误和学习模式特征，引入系统动力学理论和方法对维护人员的行为模式进行因果关系建模，突破传统认为的PSF的正交性、独立性关系，辨识潜在的维护差错动态模式，同时也提供了可集成维护人员生理参数测量以获得人因可靠性评估技术数据的模型化接口，在时域内反映了维护人员的认知、执行行为的过程机理，以维护人员的PSF和人误概率（HEP）之间因果关系分析为核心，基于Vensim/Sable develper工具平台，建立了维护人员维护差错评估仿真模型，进而依据“主导反馈回路迁移”方法分析了以培训为代表的组织性风险管控策略场景，归纳提出维护差错抑制的系统性策略建议。在此基础上，本项目紧密联系无人机系统维修组织对培训风险收益预估的需求，开发了具备软件仿真可视化界面、仿真过程动态可视的“无人机系统维修培训收益风险预估软件”并申请了软件著作权，该软件可推动维修组织和人员总结任务经验，使管理/决策者直观理解风险机理，形成对先期预判培训策略收益的“主动安全性行为”，创新性帮助我国无人机维护组织打破维护培训投入不能获得理想风险控制收益的困境，并促进无人机系统研制单位在设计中考虑重要的维护因素，全面提高无人机系统的运行安全性水平，保证公众安全和未来无人机市场的良性发展。

内容获取失败，请点击重试