Experimental Verification Research On Target Identification Errors And Information Collecting Errors Of Operator During Monitoring On Digitalized Human-interface

With the development of computer and control technology, distributed controlsystem has been changed to computer integrated digital control system gradually. Thenew features of digitalized control system have changed operator monitoring behaviorfor human-machine interface fundamentally. While the human error is the root ofaccidents in main control room, so the new control room brings great challenges tothe operator. Because of human limitations, there will happened new error modeduring monitoring in the digitalized main control room. Therefore, prediction,identification and prevention of new error mode are key contents for human reliabilityanalysis. The purpose of this paper is to ensure the accuracy and regularity of humanerror identification results, improving the validity and consistency of human erroranalysis. It can also provide clues for the optimization man-machine interface. Hence,it has positive attributions and something referential value for improve monitoringperformance of operator and human reliability in digital control system.Firstly, the author introduced the achievements and shortcomings of operatormonitoring behaviors in domestic and foreign research through literature readingmethod. And then according to the field observation summarized target recognitionerrors and information collecting errors during operator monitoring on digitalizedhuman-machine interface. The existences of sub-errors were verified by experiment.Finally, the experimental data analysis by ANOVA for repeated measurement method,correlation analysis, regression analysis and curve fitting analysis.Two target identification errors (target identification delay errors and targetidentification failure errors) of operator in the process of monitoring on digitalizedhuman-machine interface were experimental tested and measured. The simulationexperiment results showed that target recognition delay error probability is approximately equal to0.017, and the target identification failure error probability isapproximately equal to0.010. ANOVA for repeated measurement method found thesignificant difference of target recognition error probability between different fieldlocations and different eccentricity. And the error rate of target recognition increasedwith eccentricity, the fitting function is: ln()=-6.292+0.214x.Four information collecting errors (failure to collect relevant information, residualinformation collected, insufficient information collection, information collectiontimeout) of operator in the process of monitoring on digitalized human-machine wereexperimental tested and measured. And then through the simulation experiment, theexistences of these four types of errors were verified. The experimental resultsindicated that the overall error rate of target recognition is equal to0.0187, the errorrate of being unable to collect relevant information, collecting residual information,collecting information insufficient and collecting information timeout are0.0009,0.0027,0.0071and0.0080, respectively. On the basis of correlation analysis,regression analysis and curve fitting analysis, following conclusions can be made:(1)there is a significant correlation between the average reaction time of informationcollecting during operator monitoring and number of target information. And theregression equation is:=6639.87+599.311x;(2) there is a significant correlationbetween the error rate of information collecting during operator monitoring andnumber of target information. And the curve fitting function is: ln()=-6.292+0.214x.However, the error rates of four main kinds errors of information collecting are notsignificantly related to number of target information.