Research Of The Heavy Haul Freight Wagon’s Buffer Mathematical Model

With the deepening of reform and opening up, China’s economy has long maintained ahigh growth, and China became the world’s second largest economy. As the significantnational economic infrastructure, rail freight transportation has gradually been muchattention,the national railway of2011a total delivery of the goods reached3.93billion tons,the Datong-Qinhuangdao railway traffic has reached440million tons, the opening lines ofheavy freight trains is the development of rail freightan important direction. However, therapid development of heavy haul train at the same time, it exposed a series of questions,especially the larger longitudinal impulse has begun to endanger the safety of theoverloaded train, many of domestic institutions to pay special attention to the heavy trainlongitudinal dynamics research.Longitudinal impulse of the train by the impact of various factors, which hook-buffersystem as a vital component in the entire train, has a major impact on the vertical impulse.The buffer model accuracy is the accuracy of the simulation results of the longitudinaldynamics, the most important factor, so the study of the buffer model without delay.Buffer model in this paper comes from a real vehicle crash test results, by assuming thatthe buffer charactweistics, according to test results to establish the the mathematicalrelationship among the buffer parameters, the use of BP neural network to quickly andaccurately establish the HM-1buffer and HM-2buffer model, differentimpact velocity ofthe vechile crash test simulation, and compared with experimental results. To get thefollowing results: HM-1buffer model fitting of the absolute error to get the following results,an average of9.3%absolute error variance is0.0077, the absolute error of the mean at allspeeds, the maximum resistance force, the experimental data and fitting data3.3%, thelargest resistance force of the position errors are less than1mm; HM-2buffer modle fitting ofthe absolute error to get the following results, an average of7.7%absolute error varianceis0.002, the absolute error of the mean at all speeds, the maximum resistanceforce, the experimental data and fitting data7.3%, the largest resistance force of the positionerrors are less than1mm.This paper established two buffers fitting out the bumper of model of the workingprocess of the form and the curve of the test curve form high similarity; The fitting thegreatest impedance force and test the greatest impedance force size differ not quite,occurrence position is also very close. In this study, the buffer model is conducive for furtherresearch on HM-1buffer and HM-2buffer mechanical characteristics, when dealing withother type buffer model at the same time, can directly generate the test data to structure newbuffer mathematical model. This method opens up a new way for the design of overload trainbuffers and the research of overloaded train longitudinal dynamics.