Research On Asymmetric Rail Profile Optimization And Its Applicability

Heavy-duty railway transportation occupies a pivotal position in China's freight transportation methods.As the axle weight and speed of vehicles gradually increase,the problem of rail wear on small-radius curved rails becomes increasingly prominent.Therefore,related scholars can reduce the rail wear rate by optimizing the rail profile.However,when the rail profile is optimized,the line parameters cannot be guaranteed to be completely consistent with the line parameters when the rail profile is optimized.Therefore,it is necessary to carry out research on whether the optimized rail profile is applicable when the external parameters change.The research content of this paper is firstly to obtain the optimal profile under specific line parameters;secondly,to study the applicability of the optimal profile after the line parameter changes.First,the arc parameters of the rail profile are used as independent variables,and the optimal arc parameter range of the rail profile is determined by combining the asymmetric profile optimization theory and the wheel-rail contact theory.Second,the arc parameters are sampled to obtain a simulation experiment sample.In the simulation,the random impact of the wheel tread and the vehicle curve passing speed is considered,and the metal wear rate in the design cycle is taken as the target value.Finally,the independent variables and target values of the samples were trained by BP neural network to obtain an approximate model of the rail profile optimization,and then the genetic algorithm was used to solve the rail profile optimization model to obtain the optimal profile.Then,the route condition parameters were classified and based on the control variables The method is used to obtain the change law of the rail metal loss rate under the condition of a single line under the optimal profile condition.Finally,based on the Monte-Carlo simulation experiment,repeated application of the simulation experiment to determine the applicable probability of the optimal profile after facing random changes in external parameters.Based on the BP neural network-genetic algorithm,the optimal rail profile under specific track conditions was obtained.The results show that compared with the standard rail profile,the optimal rail profile has an optimization effect of reducing rail wear rate by 14.8%.Monte-Carlo simulation experiment results show that under the condition that the line condition parameters are randomly changed within the allowable range,the optimal rail profile has an 80.8% probability and has an optimization effect,of which there is a 36% probability that the optimal rail profile has an optimal effect Above 14.8%,there is a 44.8% probability that it is lower than 14.8%;there is a 19.2% probability that it no longer has an optimization effect.The results show that when the external condition parameters are inconsistent with the specific condition parameters when the rail profile is optimized in practice,the optimization effect of the optimal rail profile is likely to fall short of expectations,or even negative optimization may occur.