Load Model Of Traction Power Supply System And Its Impact Of Access To Power System

In recent years,high-speed railway has advantages of long distance transportation and ride comfort,low carbon and environmental protection has been the rapid development in our country.As a nonlinear load,it has a strong impact on the power grid,and it also shows strong randomness.The probability model of traction load can be used to study its load characteristics from the perspective of probability and statistics.Combining with probabilistic power flow algorithm,it can effectively evaluate the influence of traction load on the node voltage and the branch power flow.Firstly,a simulation model of locomotive traction power supply system and a traction load probability model are constructed.The former mainly analyzes the connection form of traction substation,the structure of traction power transmission system and the control method of converter,and builds the simulation model of locomotive,traction substation and traction network on the basis of it,and analyzes the harmonic distribution characteristics of the feeder current under different working conditions.The latter includes probabilistic model modeling and model parameter identification algorithm.The log normal distribution is used to fit the data of the feeder current of the traction network,and the harmonic current is simulated by sampling the type of locomotive and the content of the characteristic harmonic.The probability model of traction substation is described by the combination of traffic density and vehicle power.Taking the difference between the probability density data of simulated data and the probability density value as the objective function,a multi swarm particle swarm optimization algorithm is introduced to improve the accuracy of parameter identification.Secondly,the probabilistic power flow algorithm is studied.It mainly includes the comparison and analysis of several probabilistic power flow algorithms,the data sampling method of probability model,the correlation control of the input variable of the power flow equation and it's improved algorithm.It mainly introduces the core steps of the probabilistic power flow algorithm based on Monte Carlo simulation algorithm,semi invariant method and three point estimation method.And on this basis,the data sampling and correlation control problems are brought out.Considering that the probabilistic model of traction substation is a double-layer model of vehicle density and vehicle power,a Latin hypercube algorithm combined with random sampling is proposed to ensure the uniformity of the samples.Meanwhile,aiming at the sorting steps in traditional Latin hypercube sampling,a sample ranking algorithm based on simulated annealing algorithm is proposed,and the correlation information between samples is expressed in the form of objective function,and then simulated annealing algorithm is used to optimize it.The simulation results show that the method can effectively control the correlation between the input random variables.Iteffectively solves the problem that the use of traditional methods can not effectively control the correlation between the traction load.Finally,in the IEEE-14 node and IEEE-30 node test system,the changes of the node voltage and the influencing factors after traction load access are analyzed based on the three point estimation method and the improved Latin hypercube sampling probabilistic power flow algorithm.And the correctness of the algorithm is verified compared with the probabilistic power flow algorithm based on Monte Carlo simulation.The simulation results show that the probability of each node's voltage overshoot is related to the capacity and location of the traction load of the access point.Considering the correlation between input variables,the influence of branch power flow and node voltage overshoot probability is more obvious.