An online state and parameter estimation of dynamic state space model of high speed train

作者：Xie Guo^*; Zhang Dan; Hei Xinhong; Qian Fucai; Takahashi Sei; Nakamura Hideo

来源：International Journal of Control and Automation, 2016, 9(9): 51-62.

DOI：10.14257/ijca.2016.9.9.06

• Arbitrary distribution
• Bayesian theory
• High speed train (HST)
• Nonlinear dynamic characteristics
• Online state and parameter estimations
• Parameter identification methods
• Particle filter
• Posterior probability

In this paper, the state space model of high speed train is established to describe its nonlinear dynamic characteristics, whose parameters are disturbed by noise with an arbitrary distribution, and an online state and parameter identification method is proposed based on a parameters set by employing Bayesian theory and Particle Filter (PF). Firstly, the priori probabilities of all of the possible parameters are set to be equal, and the predicted states with different parameters are estimated using PF. Then the posterior probabilities of the parameters are updated by analyzing the characteristic of the measurement noise using Bayesian theory. Finally, the system sate and parameter are estimated by weighted summing all of the parameters and predicted states. The simulation results indicate that the proposed method can estimate the states and parameters of high speed train online and adaptively.

• 出版日期2016