Stochastic Control Systems with Fault Diagnosis Techniques

Stochastic control systems integrate random disturbances and uncertainties into the dynamic models that govern state evolution and output behaviour. These systems are typically described by stochastic differential or difference equations incorporating process noise, measurement noise and, in some cases, non-Gaussian disturbances. Fault diagnosis in this context encompasses the detection, isolation and estimation of anomalies or degradations in system components, followed by reconfiguration or compensation strategies to maintain performance and safety. Common approaches employ probabilistic observers or filters to generate residual signals that indicate deviations from expected behaviour, and then apply decision logic or adaptive control laws to accommodate or correct faults. Advanced methods combine probability density function (PDF) shaping, entropy-based criteria and sliding-mode or neural-network architectures to enhance robustness against heavy-tailed noise and intermittent measurement losses. The interplay of state estimation, online learning and control synthesis under uncertainty has led to a rich set of techniques for fault-tolerant control in applications such as power systems, aerospace navigation, industrial automation and networked infrastructures.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Stochastic Control Systems with Fault Diagnosis Techniques publication trend

The graph below shows the total number of articles in stochastic control systems with fault diagnosis techniques across all publications each year (not limited to Nature Index journals).

Technical terms

Stochastic control system: A dynamic system model incorporating random disturbances in its evolution and measurements.

Fault diagnosis: The process of detecting, isolating and estimating system faults using residuals or statistical indicators.

Probability density function (PDF): A function describing the likelihood distribution of a continuous random variable’s outcomes.

Residual signal: The difference between observed outputs and model-based estimates, used to indicate anomalies.

Sliding-mode control: A robust control technique that forces system trajectories onto a predefined sliding surface despite uncertainties.

Moment-generating function: A function that characterises all moments of a random variable, aiding in non-Gaussian analysis.

References

1. Non-Gaussian disturbance rejection control for multivariate stochastic systems using moment-generating function. ISA Transactions (2023).
2. Discrete‐Time 2‐Order Sliding Mode Fault‐Tolerant Tracking Control for Non‐Gaussian Nonlinear Stochastic Distribution Control Systems with Missing Measurements. Complexity (2020).
3. Two-Step Neural-Network-Based Fault Isolation for Stochastic Systems. Mathematics (2022).