摘要
在复杂人机系统中,操作员功能状态(Operator Functional States, OFS)是影响系统安全性的一个重要因素。本文基于aCAMS仿真系统实验中采集到的一系列操作员电生理信号及性能数据,采用协同进化模糊(Fuzzy Cooperative Coevolutionary, FCC)建模方法建立OFS模型,实现对OFS的评估及预测。该方法将模糊模型的主要因素,如模糊规则的前件/后件、隶属函数参数编码为不同的种群,同时优化模糊模型的结构与参数,算法中种群个体适应度函数同时考虑模型的精确性和解释性,采用分量加权求和法将多目标优化转化为单目标优化。首先将该方法用于几个UCI标准分类数据集的的建模仿真,验证该建模方法的通用有效性。然后针对于OFS建模中5个EEG和ECG指标对不同被试OFS的敏感度不同,通过Simba算法计算这些指标的重要性来进行变量选择,确定OFS模型的初始结构,再用协同进化算法对其进行结构和参数上的优化。最后的建模结果表明,协同进化模糊建模方法用于OFS的建模是有效的,优化后的模型精度在可接受范围内,所涉及的模糊规则和模糊子集总数较之优化前均有较大幅度的减少,模型结构得到了简化,模型兼具良好的精确性和可解释性。与基于GA的建模方法相比,FCC用于OFS建模展现了更加优越的性能,所建立的模型基本能反应OFS的实际变化趋势,具有一定的预测能力,可用于自适应系统中,根据模型预测结果,适时调整系统控制策略,从而实现智能化人机交互系统。
Operator Functional State (OFS) is an important factor that affect system security in complex human-machine system. This paper adopts Fuzzy Cooperative Co-evolutionary (FCC) modeling algorithm to establish OFS models based on a series of electrophysiological signals and operator performance data. This method encode the main parameters of fuzzy model in different populations, such as the antecedent/post of fuzzy rules, membership function parameters, to optimize model structure and parameters synchronously. The individual fitness function in algorithm populations consider both of model accuracy and interpretability by using a components weighted-sum method which transit multi-objective optimization into single-objective optimization. FCC method is used to establish fuzzy classification systems for five UCI data sets before OFS modeling. And the simulation results verified its effectiveness. According to the sensitivities of EEG and ECG markers of different operators are different, a Simba method is adapted to select input variables. The simulation results show that each OFS fuzzy model has good accuracy and simplified structure, evolves less rules and fuzzy sets. Comparison to GA fuzzy modeling method, FCC has better generalization performance for establishing OFS model. The final model based on the results is used to adjust control strategies, achieving intelligent human-computer interaction.
引文
[1]L.J.Skitka,K.L.Mosier,M. Burdick. Does automation bias decision-making? Int.J.Human Computer Studies.1999,51:991-1006
[2]M.Tkahashi,O.Kubo,H.Yoshikawa. Mutual Adaptive Interface:Basic Concepts. Robot and Human Communication,1993. Proceedings.2nd IEEE International Workshop on.
[3]H.Yoshikawa, O.Kubo, M.Tkahashi. Concept on Mutual Adaptive Interface and Related Experimental Study. Institute of Atomic Energy, Kyoto University Press.1994
[4]G. R. J. Hockey. Operator functional state:The Assessment and Prediction of Human Performance Degradation in Complex Tasks [M]. US:IOS Press.2003
[5]Yukiyu Kuriyagawa, Ichiro Kageyama. "A modeling of heart rate variability to estimate mental work load"-Systems, Man, and Cybernetics,1999 IEEE SMC 99 Conference Proceedings.1999 IEEE International
[6]P. Nickel, A. C. Roberts and G. R. J. Hockey. "Assessment of high risk operator functional state markers in dynamical systems-preliminary resultsand implications". in Proc. of Human Factors and Ergonomics Society Europe Chapter Annual Meeting. Turin, Italy,2005,10:26-28
[7]Jian-Hua Zhang, Xing-Yu Wang, M. Mahfouf, D.A. Linkens. Use of Heart Rate Variability Analysis for Quantitatively Assessing Operator's Mental Workload. in BMEI (1)2008:668-672
[8]M. Takahashi et al. Development of Realtime Cognitive State Estimator, Pm. of 12th World Congress of International Federation of Automatic Control.1993,7:219-222
[9]苏令波.电气操作运行人员的心理负荷研究[D].北京交通大学硕士论文.2006,12
[10]G.Conway. Effects of workload, effort and fatigue on complex performance:application and tests of compensatory.PhD Thesis, University of Leeds.2005
[11]G. F. Wilson, F. Fisher.The use of cardiac and eye blink measures to determine flight segment in F4 crews. Aviation, Space and Environmental Medicine.1991,62:959-961
[12]G. F. Wilson, F. Fisher. Cognitive task classification based upon topographic EEG data. Biological Psychology.1995,40:239-250
[13]A. Gevins, M. ESmith, H. Leong. Monitoring working memory load during computer-based tasks with EEG pattern recognition methods. Human Factors.1998,40:79-91
[14]C. A. Russell, G. F. Wilson. Air traffic controller functional state classification using neural networks. Proc of the Artificial Neural Networks in Engineering Conference. 1998,8:649-654
[15]G. F. Wilson, F. T. Eggemerier. Physiological measures of workload in multi-task environments. in Damos D. Multiple-task Performance.1991,329-360
[16]Zhang J, Mahfouf M, Linkens D.A. State-of-the-Art of Mental Workload Assessment Research. Research Report, Intelligent Systems Laboratory & Human Performance Laboratory. Department of Automatic Control and Systems Engineering, The University of Sheffield, UK.2005, 10
[17]董明清,马瑞山.脑力负荷评定指标敏感性的比较研究.航天医学与医学工程.1999,12(2):106-110
[18]C. Dussault, J. C. Jouanin, C. Y. Guezennec. EEG and ECG changes during selected flight sequences. Aviat Space Environ Med 2004.75(10):889-897
[19]C. Dussault, J. C. Jouanin, M. Philippe et al. EEG and ECG changes during simulator operation reflect mental workload and vigilance. Aviat Space Environ Med.2005, 76(4):344-351
[20]A. Gevins, M. E. Smith. Detecting transient cognitive impairment with EEG pattern recognition methods. Aviation, Space and Environmental Medicine.1999,70: 1018-1024
[21]A. R. Nikolaev, G. A. Ivanitskii, A. M. Ivanitskii.Reproducible EEG alpha-patterns in psychological task solving. Human Physiology.1998,24:261-268
[22]G. F. Wilson.Operator functional state classification using neural networks with combined psychological and performance measures. Proc of the 43rd HFES Annual Meeting.1999,1099-1102
[23]Zhang J, Mahfouf M, Linkens D.A.et al. Adaptive fuzzy model of operator functional state in human-machine system:A preliminary study. Proc of the IASTED Int.2007, 555-560
[24]Zhang J, Mahfouf M, Linkens D.A, Wang X. Adaptive fuzzy approaches to modeling operator functional states in a human-machine process control system. Int. J of Human-Computer Studies,2007
[25]刘华,张建华,王行愚.人机系统操作员功能状态的神经模糊建模方法[J].电子测量与仪器学报.2009,314-319
[26]彭小迪,张建华,王行愚,王娆芬.基于遗传算法的操作员功能状态模糊建模[J].《电子测量与仪器学报》.2009,27-30
[27]Wiegand. An Analysis of Cooperative Coevolutionary Algorithms. Ph.D dissertation. George Mason University.2003
[28]闫春红.心率变异性非线性动力学分析的研究[D].重庆大学硕士学位论文.2004
[29]Jian-Hua Zhang, A. Nassef, M. Mahfouf et al. "Modeling and analysis of HRV under physical and mental workloads". in Proc. of the 6th IFAC Symposium on Modeling and Control in Biomedical Systems. Reims, France.2006,9:189-194
[30]K.A.Greene, K.W.Bauer, GF.Wilson, C.A.Russell, S.K.Rogers and M.Kabrinsky. "Selection of psychophysiological features for classifying air traffic controller workload in neural networks". Smart Eng.Syst.Design.2000,2:315-330
[31]Hockey G.R. Wastell D.G. and Sauer J. Effects of sleep deprivation and user-interface on complex performance:A multilevel analysis of compensatory control. Human Factors.1998,40:233-253
[32]H. H. Jasper.Report of the committee on methods of clinical examination in electroencephalography [J]. Electroencephalography and Clinical Neurophysiology.vol. 10,1958:370-375
[33]GF.Wilson, F.Fisher. Cognitive task classification based upon topographic EEG data. Biological Psychology.1995,40:239-250
[34]王立新著,王迎军译.模糊系统与模糊控制[M].北京:清华大学出版社.2003,6:2-11
[35]刘静.协同进化算法及其应用研究.西安交通[D].大学博士学位论文.2004
[36]Mitchell M.An Introduction to Genetic Algorithms[M].MIT Press,Cambridge,MA.1996
[37]Hillis. Co-evolving parasites improve simulated evolution as an optimization procedure. Artificial Life, SFI Studies in the Sciences of Complexity.1991,10:313-324
[38]Paredis J. Steps towards coevolutionary classification neural networks. Proceedings of the Fourth International Workshop.1994,102-108
[39]Paredis J. Coevolutionary constraint satisfaction. Proceedings of the Third Conference on Parallel Problem Solving from Nature.1994,46-55
[40]Potter and De Jong. A cooperative co-evolutionary approach to function optimization. Proceedings of the Third Conference on Parallel Problem Solving from Nature.1994, 49-257
[41]Potter and De Jong. Evolving neural networks with collaborative species. Proceedings of the Sixth International Conference on Genetic Algorithms.1995,366-372
[42]Potter and De Jong. A cooperative co-evolutionary approach to learning sequential decision rules. Proceedings of the Sixth International Conference on Genetic Algorithms.1995,340-345
[43]Potter. The design and Analysis of a Computational Model of Cooperative Coevlution. Ph.D dissertation,George Mason University.1997
[44]Wiegand. Applying diffusion to a cooperative co-evolutionary model. See Eiben, Baeck, Schoenauer and Schwefel.1998,560-569
[45]郑浩然.基于生态特征的进化与协同研究[D].中国科学技术大学博士学位论文.2000
[46]Cao X B, Li J L, Wang X F. Research on coevolutionary optimization based on ecological cooperation.Journal of software.2001,12(4):521-528
[47]曹先彬,王煦法.基于生态竞争模型的遗传强化学习.1996,10(6):658-662
[48]曹先彬,罗文坚,王煦法.基于生态种群竞争模型的协同进化.2001,2(4):556-562
[49]Husbands and Mill. Simulated coevolution as the mechanism for emergent planning and scheduling.Proceedings of the Fourth International Conference on Genetic Algorithms.1991,264-270
[50]C.A. Pena Reyes.Coevolutionary Fuzzy Modeling.Springer-Verlag Berlin Heidelberg. LNCS.2004,3204:51-69
[51]张德丰.MATLAB模糊系统设计[M].国防工业出版社.2009
[52]C.A. Pena Reyes, Moshe Sipper. Fuzzy CoCo:A Cooperative-Coevolutionary Approach to Fuzzy Modeling. IEEE Transactions on Fuzzy Systems.VOL.9, NO.5, Oct.2001
[53]张永.基于解释性和精确性的模糊建模方法研究[D].南京理工大学博士学位论文.2006
[54]Potter, De Jong. Cooperative Coevolution:An architecture for evolving coadapted sub-components[J]. Evolutionary Computation,2000,8(1):1-29
[55]Goldberg D E, Genetic Algorithms in search, Optimization, and Machine learning. MA:Addison-Wesley Publishing Company,1989
[56]陈国良,王煦法等.遗传算法及其应用[M].北京:人民邮电出版社.2001:21-29
[57]C.J.Merz and P.M.Murphy. "ftp://ftp.ics.uci.edu/pub/machine-learning-databases/breast cancer wisconsin/".UCI repository mach.learing databases,1996
[58]Forina.M et al. "ftp://ftp.ics.uci.edu/pub/machine-learning-databases Wine/". UCI repository mach.learing databases,1991
[59]Kohavi R, John G H.Wrappers for feature subset selection [J].Artificial Intelligence. 1997,97:273-324
[60]Liu H, Motoda H. Feature selection for knowledge discovery and data mining[M]. KluwerAcademic Publishers.1998
[61]Linkens D A,Chen M Y.Input selection and partition validation for fuzzy modeling using neural network[J].Fuzzy sets and systems.1999,107:299-308
[62]Ran Gilad Bachrach, Amir Navor, Naftali Tishby. Margin based feature selection theory and algorithms [A]. Proceedings of the 21st International conference on Machine Learning[C]. Banff,Canada.2004
[63]Clark D, Schreter Z, and Adams A. A Quantitative Comparison of Dystal and Back Propagation. Australian Conference on Neural Networks (ACNN'96)
[64]Xiaoming Huo, Seoung Bum Kim, Kwok-Leung Tsui, Shuchun Wang. FBP:A frontier-based tree-pruning algorithm. Knowledge and Data Management. May 2002
[65]Brad Morantz. Automated Anomaly Detection. Georgia State University, USA.2005.5