Research On AP Deployment Optimization Method Based On CBTC System

The rapid development of Urban Rail Transit brings about the trade-off between railway capacity and wireless coverage.Both capacity and wireless coverage need to be determined by the relative position of AP,so AP deployment optimization is the core of the wireless coverage problem.DCS(Data Communication Subsystem)in CBTC(Communication-based Automatic Train Control System)can realize the two-directions communication between train and ground and provide the exact position of the train,which is the key to ensure the stability and reliability of wireless transmission in AP deployment optimization.Over-intensive AP deployment may lead to more serious co-Channel interference,lower communication system capacity and higher cost,while minimized spatial overlap may lead to discontinuous wireless signal coverage.Therefore,a set of reasonable optimization methods is particularly important when dealing with AP deployment.The topic studied in this thesis is AP deployment optimization under CBTC system.Aiming at the defects of the existing CBTC AP deployment optimization method,such as single simulation scenario,short track length,high computational complexity and low search efficiency,simulation research is carried out,and an improved optimization algorithm is given.The main work of this thesis includes:1.Based on the analysis of the existing AP deployment optimization methods of CBTC system,the AP deployment optimization method based on accurate outage probability is simulated and verified,and the relationship between shadow correlation distance and track scene is explored.The track simulation scene is extended from the tunnel to the suburbs,viaducts and cutting,and the track length is extended from the existing 300 m to 1320 m.On this basis,the extra-rail interference source is added as the third interference source,and the calculation method of accurate outage probability under three interference sources is explored.In view of the complexity of the accurate interrupt probability calculation method,the performance of three improved interrupt probability calculation methods,Wilkinson,Schwartz and CHAN,is compared through simulation research.According to the performance of these three interrupt probability calculation methods when they are applied to AP deployment optimization problem,The CHAN interrupt probability calculation method which is most suitable for the simulation environment in this thesis is applied to the subsequent simulation research.2.Aiming at the problem of cost function weight vector Change caused by simulation scene Change,the MOEA / D algorithm is used to simulate the suburban,viaduct and cutting scenes under the condition of three interference sources.The optimal weight vector of the cost function in the suburban,viaduct and cutting scenes is solved,and the simulation results are compared and analyzed with the empirical weight of the cost function in the tunnel scene.3.On the basis of the above work,aiming at the problem that the violent search method is not suitable for long-distance orbit,the genetic algorithm is used to improve the search efficiency of AP deployment optimization method.Aiming at the convergence concussion caused by the traditional genetic algorithm applied to the AP deployment optimization method,an improved elite reservation strategy is adopted,which combines the advantages of nondominated fast sorting algorithm and optimal reservation strategy to eliminate the convergence concussion and accelerate the convergence speed.Finally,the performance of violent search method and improved genetic algorithm in AP deployment optimization method is compared.