Research On Metro Traffic Scheduling Model For Man-machine Cooperation

The application of Full Automatic Operation(FAO)makes the driverless driving of trains a reality,which undoubtedly brings a qualitative leap to the development of urban rail transit.Although the unmanned train technology liberates subway drivers,it also gives dispatchers and system equipment new responsibilities and functions.The remote monitoring operation of dispatchers replaces the direct operation of the train by the driver,and the functions and structure of the system equipment are strengthened and changed.However,the transfer and allocation of human-machine functions has caused the dispatcher to increase the processing time pressure caused by the sudden increase in task volume when encountering unexpected situations,and the delay effect caused by the lack of timely and effective acquisition of accurate information or complicated operation steps under high time pressure more and more obvious.In this paper,the scheduling system of urban rail transit is taken as the research object,and the dispatcher is taken as the starting point.The work breakdown structure method,human information processing model,multi-resource theory,bayesian network and evidence theory and other models and methods are used to understand the operation of personnel.From the perspective of research on the task optimization of the dispatcher under the FAO system and the cognitive performance decline of the human-computer interaction process.The concurrency and priority operation rules of dispatchers in emergency and normal conditions are proposed to optimize the scheduling task process;the operation company should improve the operation regulations of dispatchers,optimize the work task process and increase the post staffing;the system designer should optimize the layout and display content of human-machine interface and strengthen the screening and accuracy of alarm information to improve the problem of cognitive performance degradation in human-computer interaction.Finally,combined with the analysis and consideration of the operation scheduling model of the FAO system considering the task sequence,the most fundamental optimization strategy of the subway operation scheduling system is to improve the self recovery ability of the system equipment after failure,so as to truly weaken the function and role of the dispatcher and realize the full automatic operation of the train in the full cycle and full state.The main innovations of the paper are as follows:(1)Taking the task sequence as the starting point and the driving point,the subway train dispatching model under the CBTC system and the FAO system is constructed,the problems faced by the system upgrade train dispatching system are analyzed and the research focus of this paper is further drawn.(2)The work breakdown structure method is used to decompose the basic tasks and emergency tasks of dispatchers,and the basic task and emergency task module libraries are established.The task module in the module library is decomposed along the general dimension,and a scheduling task based on cognitive activities is constructed.Combined with the multi-resource theory,a cognitive resource allocation model under emergency conditions and normal conditions is built,and further,the emergency of dispatchers is proposed.Concurrency and priority operation rules under normal and normal conditions to optimize the process of scheduling tasks.(3)Based on the theory of human information processing and the characteristics of the work tasks of the dispatchers,a cognitive processing model of the subway dispatchers based on information processing is established,and the fault behaviors of the dispatchers are identified.The interaction model of induced factors of scheduling system is constructed by m-SHEL method combined with the cognitive processing model of scheduling personnel,and five kinds of error induced factors in the interaction process are identified.(4)A simplified quantitative evaluation method for the reliability of the humancomputer interaction process of the dispatching system is proposed.Using bayesian network theory and the identification results of the dispatcher's faulty behavior and faultinduced factors,information perception,information judgment and decision-making,and information are constructed.Responding to and executing the three-stage human-machine interaction bayesian network model of the dispatching system,the model is optimized using a combination of evidence theory and node reduction algorithms,and the actual operation is carried out by taking the "Metro Dispatcher Error Recognition and Alarm Information" failure mode as an example apply analysis and propose improvement strategies.