Research On Safety Monitoring Method Of Rail Vehicle Bogie Based On RBPF

As an important part of the rail vehicle, the status of the bogie is good or bad directly affects the running security, stability and comfort. The potential security problem of bogie that used to be ignored has become currently one of the most concerned issues in the railway industry. Safety monitoring of the bogie can make the faults in the operation of the vehicle be finded and located timely, solve the problems for safe operation of the vehicle, so as to ensure safe operation of passengers and trains.Currently, safety monitoring for rail vehicle bogie mainly relies on the signal analysis methods, which needs to fix up a larger number of sensors on the vehicle and it was mainly to study mutation of detection signal characteristics, which including frequency, amplitude characteristic analysis, But the reliability of the detection result was low, so signal analysis methods have certain limitations. Parameter estimation method will be able to monitor the security status of bogie by only using a small number of sensors in practical applications, so it has certain advantages over signal analysis methods.Currently, research of parameter estimation methods is still in its infancy at home and abroad, the research has mainly focused on the parameter estimation of the normal situation and lateral system,while the vehicle failure condition and the vertical system of the parameter estimation is relative in lack.In this paper, the key theoretical and technical problems about safety monitoring of rail vehicle bogie were discussed, and the related work can be seen as follows:(1) The current domestic and international study about rail vehicle safety monitoring methods were systematic analysis.Based on analyzing basic the theory and technical characteristics of traditional Kalman Filter(KF), Extended Kalman Filter(EKF), Particle Filter(PF), Rao-Blackwellised Particle Filter(RBPF).Using two-dimensional conditional linear Gaussian state space equation, relevant parameters were estimated with EKF and RBPF, The results shows that RBPF has a better performance than EKF by comparison with a higher accuracy、faster convergence speed and so on in parameter estimation.(2) According to the actual situation of rail vehicle, the rail vehicle Multi-Body System(MBS) model was established. To ensure that the dynamic performance can be reacted veritably by MBS model, and applied into verifying RBPF parameter estimation algorithm, in this paper, the MBS model was validated and improved by two key indicators of the kinetic stability and running steadiness with the real vehicle test data.(3) According to the physical model of lateral and vertical system of the rail vehicle, the dynamical equation and state space equation of lateral and vertical system were builded by using combinatorial matrix, and observation equation was constructed in accordance with configuration scheme for actual testing sensor, The continuous state space model(which was made up of state equation and observation equation) was discretized by using accurate discretization method. By setting the fault simulation of the established MBS model and applying the parameter estimation algorithm of RBPF to estimate parameters of suspension system, the fault location and degree of judgment was realized by comparing the estimated value with the true value.(4) According to the algorithm of parameter estimation based on RBPF and the rail vehicle safety monitoring test platform, the safety monitoring software of rail vehicle bogie were developed by mixed programming with Visual Studio.NET and MATLAB, which capitalizes on the high efficiency、friendly interface of Visual Studio.NET and the strong matrix computation ability of the MATLAB, The correctness and practicability of safety monitoring method of rail vehicle bogie based on RBPF was verified through the measured test data.