Prediction-Based Distributed State Estimation Control for IoT-Enabled Train Positioning Systems

Abstract

This study addresses a prediction and distributed state estimation-based reliable control problem for IoT-enabled train positioning systems which are subject to deception attacks, missing measurements, coupling delay, process and measurement noises. To mitigate the adverse effects of sensor attacks and faults on estimation performance, a novel resilient distributed state estimation method is proposed. During abnormal measurements, a saturation mechanism with adaptive bounds is embedded into each local estimator to limit the impact of distorted measurements for preventing them from exceeding acceptable boundaries. The modified local estimates are then transmitted to the state estimation center to generate a fused estimate. Using Lyapunov stability theory and stochastic analysis, sufficient conditions are derived in the form of linear matrix inequalities to obtain estimator and control gains. Furthermore, the whale optimization algorithm combined with the distributed state estimation scheme is employed to optimally adjust the weighting values of individual state estimates, thereby reducing estimation error. As a result, the improved state estimation enhances the reliability and performance of the controller, contributing to more accurate and stable train operation. Finally, the simulation results demonstrate the efficiency of the proposed predictive estimator design under output measurements affected by diverse noises, anomalous reading, varying train operational states and external disturbances. These results confirm that the proposed approach is robust and reliable, making it suitable for practical implementation in high-speed train systems.

Keywords

• Internet of Things
• train positioning system
• prediction-based distributed filtering
• physical faults
• cyber-attacks
• whale optimization algorithm