Automatic control system for pneumatic braking of a freight train with braker parameter identification function

A mathematical model has been developed for an adaptive automatic brake control system for a freight train using an Extended Kalman Filter (EKF) to identify the actual parameters of the train and its braking system. To evaluate the performance of the proposed automatic control system, a simulation experiment was carried out using a probabilistic formulation of the research problem and the developed automatic brake control model. The purpose of the experiment is to determine the probabilistic characteristics of the random variable representing the deviation of the train’s stopping coordinate from the specified value. The adequacy of the number of simulation runs for obtaining reliable statistical characteristics in this study was determined using a method for estimating the confidence interval bounds for the sample mean and variance for the given number of trials. Histograms were obtained for the distribution of stopping-point deviation during targeted braking with the EKF algorithm switched off and on. To estimate the probability that the stopping-point deviation exceeds the specified limit, a theoretical distribution law for the random variable was selected and verified using the nw2 goodness-of-fit test. It is shown that without the proposed algorithm for identifying braking system and train parameters (the Extended Kalman Filter, EKF), the deviation from the target stopping point during precision braking can lead to serious violations of train-operation safety. The proposed automatic pneumatic brake control system provides high-quality control, improving the accuracy of targeted stopping of a train at a specified track coordinate. The work was carried out at the expense of budget financing within the framework of the state assignment No. 103-00001-25-02 dated 20.03.2025.