Research On Redundancy Technology And Application Of Computer Interlocking System

The computer interlocking system maintains long-term safety and reliability plays a significant role in guaranteeing the stable operation of the railway station.As redundancy technology is the main method to improve the system reliability,thus research on redundancy technologies of the computer interlocking system is of great meaning.Among the mainstream redundant structures,double 2-vote-2is widely used in domestic and international computer interlocking systems relying on its high reliability and high safety,however,the core technologies of which are monopolized by the abroad,domestic study of maintenance strategies,redundant communication and interlocking software redundancy based on double 2-vote-2 is far from sufficient.In order to achieve higher reliability and safety,this paper focuses on the research of computer interlocking system based on double 2-vote-2.Two kinds of optimization strategies areproposed based on double 2-vote-2 maintenance strategy.Under the first strategy: the subsystem will keep on working if there exists at least one normal module when single-module detectable fault occurs,which sacrifices a small part of safety performance to obtain higher reliability meanwhile ensuring the normal operation of the system.Under the second strategy: the remaining module of prime keeps on working instead of prime-slave switching when single-module detectable fault of prime occurs while slave is still in normal state,which can improve the system reliability while reducing the switching frequencies to avoid switch oscillation caused by frequent switching.The simulation results show that the optimized double 2-vote-2 has better redundancy performance.In order to solve the problems of transmission delay,packets dropping and network congestion caused by the surging interlocking data on the network,using optical fiber loop-to-loop network to set up a distributed transmission network which achieves the self-diagnosing and self-repairing aiming at the network disconnection of single-loop and double-loop,as a result of which obtains higher reliability in data transmission than traditional redundant communication.On the above basis,proposing a kind of distributed data synchronous protocol named as UDS(UDP-based Data Synchronization Protocol),in which the data synchronousperiods will be divided into several time slots and assign them to the sending task and resending task respectively both of which will execute with Round-Robin by time slots and the receiver will periodically send retransmission acknowledgement to the sender.Through theoretical derivation,the good effect on controlling the network congestion by UDS is proved.The analysis of the experimental data verifies the improvement of network congestion control under UDS protocol.According to the unreliable problems exists in fault diagnosing,prime-slave synchronization and switching of interlocking software,combining the optimized double 2-vote-2 and redundant communication,designs a double 2-vote-2 redundant management simulation system,this system establishes mechanism of fault diagnosing and repairing,moreover,uses mechanism of state detection to ensure accurate states synchronization between prime and slave,guarantees smooth transition in prime-slave switchingbased on model of system redundancy migration.Software testing simulation platform is established under experimental environment to test the redundancy management system,results show that the system realizes multilevel redundancy upon data,processes and interlocking system,improves the redundant performance of the computer interlocking system.