Performance Parameters Prediction Design Of Freight Train Braking System Based On BP Neural Network

Nowadays,heavy freight loading,the main phenomenon of railway,is the dominant research direction of railway transportation.And the performance of train's braking system is a very important part for the development of heavy-load transportation.Normally,the traditional braking system must be tested by circuit and set equipment.It often takes time and funds but the unsatisfactory result.Hence,the vital problem focus on the how to design more robust braking system test platform.In light of the above issue,this thesis establishes the gas flow state model of operating brake system.And analysis the gas state equation,mass flow equation and mechanical movement of the gas chambers.Using neural algorithm to predict the parameters of the model and comparing the gap between the predicted data and original data.Finally,small marshalling braking system of heavy-loading trains are used to verify this scheme.Firstly,this thesis introduces the structure of 120-type distribution valve,the pressure relationship between the cylinder chambers and the gas flow characteristic to analysis each stage of the braking process.Secondly,establishing a mathematical model of each cylinder chamber of truck brake system to get the characteristics of the fluid in the inflated and deflated process.And the simulation can be divided into three modules: initial inflation conditions,common barking conditions,and acceleration and mitigation conditions.The model laid the foundation for the establishment of the mathematical model of the 120-type air control valve.Using BP neural network algorithm effective throttling coefficient of prediction model to analysis the value of parameters for each vehicle position under different conditions.Then,the conclusion can be drawn by the experimental data of the prediction model and the original data.Finally,the software platform of the host computer for establishing the test platform is thoroughly discussed.And bring the braking simulation and test platform for heavy-loading trains of small marshalling vehicles.