Networked Control Systems and State Estimation Techniques

Networked Control Systems (NCS) integrate sensors, actuators and controllers via shared communication networks, enabling real-time monitoring and control across spatially distributed components. The interconnection brings benefits of flexibility, scalability and reduced wiring costs, but also introduces variable latency, packet loss and bandwidth constraints. Accurate state estimation—the reconstruction of unmeasured internal variables from noisy or sporadic measurements—is fundamental to maintain stability and performance. Classical approaches such as Kalman filtering assume periodic data delivery, but are challenged by resource limitations. Recent trends emphasise co-design of control and communication layers, exploiting model-based prediction and adaptive sampling to balance estimation quality with network load. Event-triggered mechanisms reduce unnecessary transmissions by sending updates only when a measure of system variability exceeds a threshold. Age-of-Information (AoI) metrics further refine scheduling by accounting for the freshness of data at the controller. Together, these advances promise robust, efficient operation in applications ranging from industrial automation and smart grids to autonomous vehicles and the Internet of Things.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Research from all publishers

Dynamic event-triggered control and estimation frameworks have been developed to enhance resource efficiency in NCS by introducing auxiliary dynamic variables and adaptive threshold parameters. These schemes systematically trade off communication usage against estimation accuracy, demonstrating improved performance over static time-triggered strategies in simulation and real-world benchmarks. The fundamental construct unites controller design with event criteria, offering flexible tuning for diverse system requirements.

The send-on-delta sampling paradigm, a signal-dependent temporal sampling method, has been established as a core strategy for sensor networks within control loops. By triggering transmissions only when sensor readings deviate beyond a predefined increment, it achieves substantial reductions in network traffic while bounding estimation error. Its analytical evaluation on bandlimited signals reveals effectiveness guarantees and practical guidelines for threshold selection in wireless and industrial settings.

Machine learning techniques have been integrated with state estimation using two-way Gaussian process regression and Age-of-Information-aware scheduling. In this co-design approach, missing state or control updates are predicted locally with quantified uncertainty, while communication resources are allocated to minimise average AoI under reliability and power constraints. Results indicate the method can double the number of controllable actuators and significantly outperform traditional event-triggered and round-robin baselines.

Networked Control Systems and State Estimation Techniques publication trend

The graph below shows the total number of articles in networked control systems and state estimation techniques across all publications each year (not limited to Nature Index journals).

Technical terms

Networked Control System: A control architecture in which sensors, controllers and actuators communicate over a shared network, often subject to latency and packet loss.

State Estimation: The process of inferring unmeasured internal variables of a dynamical system from available measurements and a mathematical model.

Event-triggered Mechanism: A communication strategy that initiates data transmission only when a predefined condition, based on system state or error, is met.

Send-on-Delta: A sampling method in which sensor data are transmitted only when the change from the last sent value exceeds a specified threshold.

Age of Information (AoI): A metric quantifying the time elapsed since the generation of the most recently received update, used to assess data freshness.

References

1. Dynamic Event-triggered Control and Estimation: A Survey. Machine Intelligence Research (2021).
2. Send-On-Delta Concept: An Event-Based Data Reporting Strategy. Sensors (2006).
3. Predictive Control and Communication Co-Design via Two-Way Gaussian Process Regression and AoI-Aware Scheduling. IEEE Transactions on Communications (2021).