Advanced Controls and Cyber Physical Systems

Advanced Controls

Advanced controls move beyond linear and steady state approaches to dynamic process modeling, representing physical systems and processes with a combination of simple dynamic model elements and high-fidelity models for more complex system components. These high-fidelity models are developed, then reduced (simplified) and configured to run in real time (with time scales on the order of milliseconds) to represent the dynamics of the complex component within the system. Using reduced, fast running models in conjunction with estimation algorithms and other types of predictive algorithms, an overall control solution can be derived to enable model-based control for real-time processes.

These advances within the Sensors and Controls program enable control with fast dynamics for non-steady-state operation and inherently nonlinear systems.

Cyber Physical Systems

Energy and carbon management systems of the future will undoubtedly integrate multiple renewable power systems, energy storage systems, carbon capture (in order to supply stable power to the electric grid with greater efficiency), net-zero greenhouse gas emissions, and favorable economics. As these systems become more tightly integrated, the challenges faced by their control systems increase in complexity.

NETL has ongoing R&D focused on developing advanced control algorithms to meet the performance challenges of hybrid power systems that feature multiple energy components, carbon reduction technologies, and other assets. NETL uses hybrid virtual and physical—that is, cyber-physical—systems to develop and test advanced sensor and control technologies. The HYPER facility at NETL leverages software models integrated with operating hardware as a platform for testing and design of advanced energy system components, control method development, integration methods, and optimized sensor placement approaches.

Inherent in the operation of cyber-physical systems is the need for models that can respond to external stimuli (including physical, virtual, and combinations), sensors, actuators, and also contend with system behavior such as software scheduling and communication delays. A compounding factor on top of this high-level complexity is the need for models that operate at fast time scales approaching real time for complex components while maintaining sufficient physical fidelity, for which existing methodologies are lacking.

Meeting these challenges requires significantly advanced control compared with that offered by traditional proportional-integral-derivative (PID) control and needs to be more robust than linear model predictive control algorithms. To this end, NETL has supported extensive R&D in the areas of advanced controls, system identification, and real-time modeling.