Research On Optimization Of Communication Mode On Train Communication Network And Reliability Analysis

With the development of rail transportation in our country, the train control network system based on train communication network (TCN) subsititutes the traditional centralized control system, and is widely used in high-speed train. As one of the key techonologies of modern high-speed train, train communication network is an important factor of safety and normal operation of train and comfort of passagers. While modern high-speed train is developing more and more automatically, comfortably and high-speedly, the devices on train will become more diverse;their functions will becomplex.As a result, the types of data on train communication network will also become more diverse, and the amount of data will continue to grow; therefore, the train control will become increasing complex in the future. Hence, a lot of researchers focus on opatimizing the train communication network used in high-speed train, studying new types of train communication network and analyzing the reliability of train control network.The detailed research work and contribution are shown in follows:1)Optimizing messagecommunicationon train communication network. This paper discusses the periodic and aperiodic message communication procedure on train communication based on TCN protocol. For periodic message, this paper optimizes the periodic polling table using a swarm optimization algorithm with adaptive mutation. The main principles of optimization are overall balance and individual diversity; it will try to improving the real-time responsiveness of aperiodic message transmission while keeping the high bus utilization ratio for periodic message transmission. For aperiodic message, this paper optimizes the event arbitration using swarm optimization algorithm; there are two methods adopted in optimization, the one is assigning device addresses reasonably and the other is extending the priority levels of aperiodic message. Because the algorithm is running offline, then the output configuration is put on train communication network, it will not affect normal real-time operation of train.2) Research on train communication network based on the switched industrial Ethernet. This paper disusses the MVB using the model of industrial Ethenet switch, and illustrates the disadvantages of low efficiency and poor concurrence for MVB on data transmission. This paper displays a switched Ethernet with ring topology to discuss the communication delay of network under different network structures and data traffic; the upper bound of the communciaiton delay is given; the weighted fair queueing (WFQ) scheduling algorithm is adopted for queue scheduling in switch. The result shows that it is feasible to using industrial Ethenert to substitute MVB. This paper also considers the distribution of devices on switches of Train Communication Network based on switched Ethernet, and proposed anlmmunity Particle Swarm Optimization (IPSO) to solve problem with balancing the load on switches and decreasing the communication delay of important real-time data flow.This paper also makes a research on optimization of scheduling algorithm, and proposes a priority-based WFQ scheduling algorithm.3) Analyzing the reliability of train network control system and making a further research on system reliability model. Train network control system mainly adopts redundancy techonology to increase the reliability of the system. Train network control system consists of a series of k-out-of-n systems using reliability block diagram, and the redundancy of master devices is equivalent to an equal load-sharing k-out-of-n system. Hence, it is easy to derive the reliability of the system. Because of wide application of k-out-of-n systems, this paper makes further research on the system, and the reliability and operational scheduling optimization of equal load-sharing k-out-of-n phased-mission systems are studied.