Online quantitative safety monitoring approach for unattended train operation system considering stochastic factors

Online safety monitoring is the key technology to the realize unattended train operation (UTO). So, online quantitative safety monitoring method is proposed to solve the state space explosion problem of the traditional model checking method. The quantitative safety level is defined to quantitatively describe the safety level of the operational state of UTO. To begin with, the composite transition graph of the linear hybrid automata (LHA) of train tracking control and the probabilistic hybrid automata (PHA) model of moving block control principles is constructed based on the composition rules between hybrid automata. Then, the reachable probability distribution of dangerous states can be obtained by verifying the established transition graph with abundant simulation results. Furthermore, the safety constrained boundary of the selected stochastic parameters in bounded time can be achieved for the corresponding quantitative safety level by using the proposed Safety Constraint Computation Algorithm. Finally, based on the performances of stochastic events evaluated by hybrid automata online, the safety status of UTO can be quantitatively monitored in real-time.