傅山　王黎靜　黃丹

摘要：本研究圍繞“人在環(huán)復(fù)雜系統(tǒng)的控制問題”這一核心科學(xué)問題，從人機(jī)不同形式智能體融合及其相互影響機(jī)理入手，構(gòu)建復(fù)雜智能系統(tǒng)的數(shù)學(xué)模型并通過科學(xué)的實(shí)驗(yàn)及觀測(cè)方法對(duì)模型進(jìn)行驗(yàn)證，最終確定用于民機(jī)駕駛艙人機(jī)工效評(píng)價(jià)的指標(biāo)體系和綜合評(píng)價(jià)方法。本年度的主要研究進(jìn)展從以下四個(gè)方面展開： 1.基于人在環(huán)復(fù)雜系統(tǒng)的人機(jī)耦合策略模型。從通過信息感知、信息分析和決策以及操控策略執(zhí)行三個(gè)模塊的具體設(shè)計(jì)與實(shí)現(xiàn)來完成人-機(jī)-環(huán)境復(fù)雜系統(tǒng)模型對(duì)于飛行員系統(tǒng)模型的設(shè)計(jì)要求，構(gòu)成了飛行員系統(tǒng)模型的設(shè)計(jì)與實(shí)現(xiàn)，最終形成了基于規(guī)則的人機(jī)耦合策略模型。在此基礎(chǔ)上，融入人為因素影響因子體系，深化飛行員與飛機(jī)系統(tǒng)及飛行環(huán)境系統(tǒng)的交互從而將人為因素的研究具體化到-機(jī)-環(huán)境系統(tǒng)仿真中的每一個(gè)環(huán)節(jié)中去。 2.多維駕駛艙人機(jī)工效綜合評(píng)價(jià)體系。通過構(gòu)建具有模塊化分布式構(gòu)架的機(jī)組工作量測(cè)量與評(píng)估仿真實(shí)驗(yàn)平臺(tái)，進(jìn)行了機(jī)組工作量測(cè)量的行為與生理參數(shù)集成采集，以復(fù)雜系統(tǒng)理論為依據(jù)，構(gòu)建分層次多維度人機(jī)工效綜合評(píng)價(jià)指標(biāo)體系，可以清晰的體現(xiàn)出飛行員在飛行任務(wù)中各個(gè)時(shí)刻在各方面的表現(xiàn)，可以對(duì)飛行績(jī)效的降低提供具有診斷性的結(jié)果，為駕駛艙設(shè)計(jì)提供參考。 3.認(rèn)知行為信號(hào)處理與模式識(shí)別。以駕駛艙設(shè)計(jì)為背景和出發(fā)點(diǎn)，通過對(duì)駕駛艙設(shè)計(jì)理念和布局規(guī)則的深入理解，針對(duì)認(rèn)知過程產(chǎn)生的生理信號(hào)與狀態(tài)量，結(jié)合人因?qū)W理論以及特定的飛行操控，將飛行員的肌電信號(hào)特征作為研究對(duì)象，通過信號(hào)的分解、相關(guān)性分析、選擇、組合四個(gè)過程，提出基于EEMD 和Hilbert 變換的動(dòng)態(tài)疲勞評(píng)價(jià)方法。結(jié)果顯示基于平均瞬時(shí)頻率的疲勞指標(biāo)能有效的表征動(dòng)態(tài)肌電信號(hào)的疲勞趨勢(shì)。方法可以推廣形成一系列行之有效的信號(hào)處理體系，最終為駕駛艙的設(shè)計(jì)提供指導(dǎo)或參考，以及為駕駛艙適航符合性驗(yàn)證提供幫助。 4.機(jī)組工作負(fù)荷綜合評(píng)價(jià)體系與預(yù)測(cè)模型。通過對(duì)機(jī)組工作負(fù)荷影響因素研究，建立時(shí)間壓力對(duì)工作負(fù)荷的影響模型。確定了時(shí)間壓力0.8為工作負(fù)荷“redline”，為機(jī)組工作負(fù)荷預(yù)測(cè)模型中的時(shí)間線分析的臨界負(fù)荷值確定提供了理論基礎(chǔ)。在此基礎(chǔ)上，通過感知任務(wù)、認(rèn)知認(rèn)務(wù)、單任務(wù)、雙任務(wù)、多任務(wù)操作實(shí)驗(yàn)，確定它們之間的關(guān)系模型，為機(jī)組工作負(fù)荷預(yù)測(cè)模型中的負(fù)荷計(jì)算部分奠定實(shí)驗(yàn)數(shù)據(jù)理論基礎(chǔ)。

關(guān)鍵詞：人機(jī)工效；復(fù)雜系統(tǒng)；人機(jī)耦合策略模型；多維評(píng)價(jià)模型；經(jīng)驗(yàn)?zāi)B(tài)分解；希爾伯特黃變換

Abstract：The project aims at building a mathematical model for complex intelligent systems to integrate pilot and automatic flying system. It will also lead to the establishment of a comprehensive evaluation techniques and systematic methods for the ergonomics in the civil aircraft flight deck. The main progress during this year is described in the following four aspects： 1.Human-machine coupling strategy model based on man-in-loop complex systems. The specific design and implementation， through information awareness， information analysis and decision-making， and implementation of control strategies， accomplished the requirements of the complex man-machine-environment model on the pilot system model， which constituted a model for design and implementation of the pilot system， and ultimately formed a human-machine coupled rule-based strategy model. In addition， the integration of human factors is also incorporated for a more completed man-machine-environment system simulation. 2.Multidimensional cockpit ergonomics evaluation system. By building a simulation platform for crew workload measurement and evaluation with modular distributed architecture， integrated collection for behavior and physiological parameters of crew workload measurement were implemented. Based on the complex system theory， a hierarchical multidimensional comprehensive evaluation index system for ergonomics was built to provide diagnostic results to the reduction of flight performance. It also set a reference for the cockpit design. 3.Signal processing and pattern recognition for cognitive behavior. The pilot EMG signal characteristics were taken as the research object aiming at the physiological signal and states during the cognitive processes. A dynamic fatigue evaluation method based on the EEMD and Hilbert transformation was proposed. The results showed that the fatigue index based on the average instantaneous frequency can effectively characterize the fatigue trend of the dynamic EMG signal. 4.Comprehensive evaluation system and forecasting model for the crew workload. Through the study of the influencing factors on the crew workload， an impact model of the time pressure on the workload was established. Furthermore， operation experiments of perception task， cognitive task， single task， dual-task， and multi-task， the relationship model between them was determined， and theoretical basis of experimental data for the load calculation part in the forecasting model of crew workload was laid.

Keywords：Ergonomics； Complex Systems；Human-machine coupled modeling； Multi-dimension Evaluation Model； Empirical