Real-time Performance Analysis And Optimization Of Urban Rail Trains Safety Monitoring Network

Urban rail transit safety monitoring network can acquire trains operation information in real time, and it can provide an effective quality assurance to ensure train operation safety. This paper takes ubran rail train safety monitoring network based on Ethernet as main object of study, and analyses the features and data transmission delay of the monitoring network system. In order to further improve the real-time performance, the optimization method of network structure is proposed. The main research contents of this paper are illustrated in the following aspects:(1) This paper systematically analyses structure features and functions of urban rail transit safety monitoring network and then divide network structure into monitoring layer, access layer, transport layer which interaction through sense data. It classifies and prioritize the data in network according to their features and importance. The data queue scheduling policies, data transport protocol and implementation of data priority are studied to meet the need of real time data transfer, which lays the hardware and software foundation for network real-time property.(2) Build up the ubran train safety monitoring network real-time analysis model. Firstly, this paper analyses the relationship between real-time and delay of network transmission, and it uses functional relation formula delay to quantitative description this relationship. Then, an average transmission time delay model of data queue with priority is established by using knowledge of the queueing theory and viewing data in each coaches as a M/M/1 model with priority. Finally using network calculus theory, the model of data transmission delay upper bound in monitoring network is built and it can get data transmission delay in the worst case.(3) Build up the ubran train safety monitoring network real-time optimization model. The network optimization problem is divided into the physical topology optimization and the logical topology optimization, according to the practical usages. Physical topology optimization model aims to design a network structure with highest reliability, considering the construction costs and structure symmetry. It is solved by PSO algorithm. After that, logical topology optimization model aims to minimize the whole network delay by choosing optimal data transmission path for data in different train coaches, and it is solved by the shortest spanning tree algorithm.(4) Using a concrete example to verify the effectivenessn of network real-time anlysis and optimization model proposed in this paper. Firstly, this paper illustrates how to calculate average delay and maximum delay using network real-time analysis models, and compare with OPNET simulation result, verifying the validity of the model. At the same time, the analysis of the example indicates that urban rail transit safety monitoring network based on Ethernet has a good real-time performance, and it can meet the requirements of the monitoring network. On that basis, this paper uses network real-time optimization model to optimize train monitoring network structure, and the optimization results show that the network performance of real-time and reliability is improved significantly compared with previous network structure by using network calculus real-time analysis model.