Da Xue

Actuator fault-tolerant control of networked distributed processes with event-triggered sensor-controller communication

This paper focuses on the problem of fault-tolerant control of spatially distributed systems modeled by parabolic PDEs with networked sensors and actuators and actuator faults. Based on a suitable finite-dimensional model that captures the dominant dynamics of the infinite-dimensional system, an event-triggered networked control system that enforces closed-loop stability with minimal sensor-controller communication is initially designed. The model is used by the controller to generate the necessary control action when communication is suspended, and its state is updated using the real-time measurements when communication is restored. Communication is triggered when a state-dependent threshold on the model estimation error is breached. The communication threshold is obtained using Lyapunov techniques and is explicitly characterized in terms of the fault magnitude, the model and controller design parameters, and the actuator locations. This characterization captures the potential impact of faults on both closed-loop stability and network resource utilization, and is used to devise stability-based and performance-based fault accommodation strategies that guarantee closed-loop stability and keep network resource utilization to a minimum in the presence of faults. Finally, the results are illustrated using a diffusion-reaction process example.

Output feedback-based event-triggered control of distributed processes with communication constraints

This work presents a methodology for the design and analysis of an output feedback-based event-triggered networked control system for spatially-distributed processes with uncertain low-order dynamics, a finite number of spatially-distributed output measurements and sensor-controller communication constraints. Based on an approximate finite-dimensional system that captures the infinite-dimensional systems dominant dynamics, a model-based controller is initially designed and implemented. A state observer, with well-characterized state estimation error convergence properties, is also designed and embedded in the sensor to estimate the slow states of the infinite-dimensional system based on the available output measurements. The observer state is used to update the model state at times when sensor-controller communication is permitted. The update is triggered whenever a certain stability threshold on the model estimation error is breached. The threshold is explicitly characterized in terms of the controller and observer design parameters, the model uncertainty and the control actuator and measurement sensor locations. The connections between the developed output-feedback-based event-triggered control approach and the full-state feedback-based strategy are identified and discussed. Finally, the proposed methodology is illustrated using a simulation example.

Integrating Model Identification and Model-Based Control of Networked Process Systems

Abstract In this work, we present a methodology for the integration of model-based control and model identification for networked process systems subject to communication constraints. The methodology aims to enhance the stability and performance properties of the networked closed-loop system in the presence of process parameter variations while simultaneously reducing the overall rate of information transfer between the plant control subsystems. This is accomplished by incorporating within the time-triggered model-based control strategy an event-triggered model parameter update scheme that keeps the plant-model mismatch to a minimum and averts the need for increasing the communication rate in response to the destabilizing influence of process drift. A process monitoring scheme with an instability alarm is devised to trigger model identification whenever a certain instability threshold is breached. The results are demonstrated using a chemical process example.

Analysis and accommodation of communication failures in event-triggered networked distributed control systems

This work presents an approach for the analysis and accommodation of sensor-controller communication failures in spatially-distributed processes controlled over a resource-constrained communication medium. We focus on distributed processes whose dominant dynamics are finite-dimensional, but uncertain, with a finite number of spatially-distributed networked control actuators and measurement sensors. Initially, a model-based networked output-feedback controller is designed based on an approximate low-order model that captures the slow dynamics of the infinite-dimensional system. The controller is implemented using an event-triggered sensor-controller communication policy, which determines the times when the sensor data need to be transmitted to the controller to update the model state based on a suitably-designed stability threshold. An assessment of the networked control systems robustness to failures in the communication medium is carried out using an online forecasting approach that tracks the growth of the estimation errors during times of communication outages. An explicit characterization of the allowable down time that communication losses can be tolerated without compromising closed-loop stability is obtained in terms of the various control system design parameters, and used to devise various accommodation measures that enhance the networked closed-loop systems robustness against communication losses. A simulated diffusion-reaction process example is used to illustrate the developed approach.

Resource-aware fault accommodation in spatially-distributed processes with sampled-data networked control systems

This paper addresses the problem of fault-tolerant stabilization of spatially-distributed systems with uncertain low-order dynamics, sensor-controller communication constraints, discretely-sampled state measurements and control actuator faults. An approximate finite-dimensional model that describes the dominant dynamics of the infinite-dimensional system is initially used to design a model-based event-triggered networked control system with a well-characterized stability-based communication-triggering threshold under the assumption that state measurements are continuously sampled. A modification of the communication threshold that takes into account the systems limited ability to monitor the state under discrete measurement sampling is then developed. Based on a forecast of the evolution of the model estimation error, an upper bound on the error growth over each sampling interval is obtained and to derive a tighter communication trigger that is used for sampled-data implementation. The modified threshold is explicitly characterized in terms of the fault magnitude, the model and controller design parameters, and the control actuator locations. Based on this characterization, resource-aware fault accommodation strategies are devised to achieve the desired balance between the fault-tolerant stabilization and reduced network utilization objectives. Finally, the results are illustrated using a diffusion-reaction process example.

Supervisory logic for control of networked process systems with event-based communication

This paper presents a methodology for supervisory event-based control of networked process systems subject to uncertain dynamics and communication constraints. The proposed methodology aims to balance the overall plant stability with the local performance needs of the component subsystems, while simultaneously optimizing the extent of information transfer between the distributed control systems. Initially, a quasi-decentralized networked control structure is developed consisting of a set of local model-based controllers that communicate with each other over a shared network at discrete times. Using Lyapunov techniques, appropriate thresholds on both the local model estimation errors and the local control Lyapunov functions are then derived and used as triggers for communication with a higher-level supervisor. In the event that either threshold is breached, an alarm signal is sent by the local controller to the supervisor which keeps track of all the alarm signals reported by the various units and then decides which subsystems should communicate with one another. The supervisory logic is designed to take the behavior of both the local and composite control Lyapunov functions into account, in order to meet both the stability and local performance requirements. Finally, the proposed methodology is illustrated using a networked chemical process example.