This paper discusses adaptive identification and control (AID&C) techniques to enable automated online identification and control of SMRs. Adaptive system ID allows engineers to rapidly measure system dynamics, calibrate sensors channels, determine loop processes, and quantify actuator authority for the various reactor control loops. Adaptive control can automatically tune these loops and adjust plant processes to optimize conditions for peak performance and power production. Another advantage of the adaptive ID and control approach is that these tools can be used during reactor operation to monitor active and passive components. Adaptive system ID techniques are used to measure loop-transfer characteristics. Presented is a practical approach that uses adaptive model-matching tools to identify the coprime factors of the local loops. This has the advantage over model based approaches since coprime factors can be identified on the real system using real data. Adaptive control enables auto-tuning of controller parameters to meet operational specifications. Using the coprime factors, all controllers that stabilize the plant can be parametrized by a free Q-parameter that can be changed to meet control system objectives and improve performance, and the tuning is performed using adaptive techniques. The controller architecture presented provides several desirable and necessary features: e.g., a default fail-safe mode of operation, stability in the presence of communications failures, guaranteed stability, and robustness. An advantage of the adaptive structure presented here is that control system stability can be guaranteed, even during adaptation by ensuring certain norm conditions on the Q-parameter and estimated plant uncertainty. More importantly, the Q-parameter can be monitored during operation, providing a real-time estimate of the changes in the plant resulting from changes in the reactor itself. This signal monitors the dynamics of each loop, providing information about the reactor from the perspective of the control process. Online monitoring using AID&C can be used to better track control system transients that result in reactor trip, thus avoiding undesirable reactor trips and diversion events. And, there is a potential that the system can better adapt to changing operating conditions during plant transients including load following procedures.
Joint structure for the real-time estimation and control of automotive dry clutch engagement
