control system

Control Co-Design

Concurrent Control Engineering for Optimal System Design

ARPA-E Program Director Dr. Mario Garcia-Sanz discusses his background and upcoming workshop.  

You are leading an upcoming workshop entitled “Control Co-Design for Wind and Marine-hydro-kinetic systems” – what is “control co-design”?

Control engineering is the application of mathematics, physics and technology towards autonomous control of physical systems. Control engineers take data about system status and performance, and use microprocessors, algorithms, circuits and actuators to improve system conditions and, ultimately, regulate variables automatically. The system can include mechanical and electrical components, chemical and biological characteristics, thermodynamics and fluid dynamics, aero- and hydro-dynamics, network interactions, and more.

control system

Fig.1. Control system

Control engineering is not limited to finding new ways to regulate existing systems. It can be used to design an entire system from the ground up. Instead of classical design method, where each engineering team (mechanical, electrical, electronics, control…) is an independent step in a sequential process, control co-design brings together various technical disciplines to work concurrently from the start. This methodology enables a more optimal design—with better system dynamics and controllability, among other advantages – that often results in lower system cost and improved reliability.

Why is ARPA-E interested in control engineering? What problem are you trying to solve?

ARPA-E looks for innovative approaches that “change what’s possible” by breaking through existing conventions and limitations. Control co-design empowers control engineers open new possibilities using its alternative design philosophy.

Current design process imposes design constraints at every step.  Incorporating control principles and capabilities from the beginning, the designer avoids imposing unnecessary and or sub-optimal design constraints on the final system.

 (a) Classical sequential design process vs. (b) Control co-design

Fig.2. (a) Classical sequential design process vs. (b) Control co-design

In particular, systems that require multiple disciplines must simultaneously take into account all the smaller systems they’re made of, which is particularly difficult when system dynamics are involved. Control co-design techniques consider subsystem interactions from the very beginning of the design, and proposes solutions that may not be otherwise achievable. 

There are many ways to apply these control ideas to system designs in a concurrent manner including direct application of control principles, co-optimization, and co-simulation. Any energy system that contains dynamics is a good candidate for optimization through control co-design, with many potential applications within ARPA-E’s interest area! 

Some examples of potential candidates for control co-design optimization include:

 Electro-mechanical systems (e.g., onshore and offshore wind turbines, tidal and wave energy converters, airborne wind energy systems and electric airplanes)
• Networks (e.g., electrical power systems with distributed generation and storage systems, traffic control grids, wind farms, wave farms and tidal farms)
• Water and bio systems (e.g., wastewater treatment plants, desalination systems and digesters)
• Energy storage systems (e.g., batteries and flywheels)
• Thermal systems (e.g., buildings and HVAC systems) 

Garcia-Sanz Figure 3

Fig.3. Potential areas to apply control co-design

Why the particular workshop focus on wind and marine-hydro-kinetic systems?

These systems are excellent candidates for control co-design optimization, because they contain significant dynamic characteristics and include the interaction of mechanical, electrical, aerodynamics and hydrodynamics subsystems.

Our initial calculations show that applying control co-design principles to wind or marine-hydro-kinetic systems could result in an over 35% decrease in levelized cost of energy. A major objective of the workshop is to refine the key assumptions made in this initial analysis. We believe the energy savings could be remarkable, and we look forward to additional feedback.

How does your background relate to control led system design? When did you develop a passion for this approach?  

Control engineering truly is my passion. Over the course of my career, I have always looked for opportunities to bridge the gap between new control theory and advanced real-world applications using a concurrent engineering philosophy. As a university professor, consultant, and entrepreneur I have had the opportunity to design robust control solutions for many companies and space agencies over the last 30 years. I am so grateful to see many of these solutions included in some of my publications, implemented in commercial wind turbines, space satellites or water treatment plants, or developed further by some of my former students.

One of my most memorable experiences working in this field is the direct-drive, variable-speed, pitch-controlled 1.65MW wind turbine we designed at the end of the nineties. It eventually became one of the largest wind turbines in the market. The machine didn’t need a gearbox and incorporated a full-power converter which allowed us to control the aerodynamic efficiency and the grid variables simultaneously and independently. By applying control co-design concepts to the pitch control system, we produced a very smooth and robust rotor speed control system, and reduced significantly the tower vibration and the corresponding mechanical fatigue of the system. This enabled us to introduce a machine with a tower that cost nearly 30% less than the immediate competitor, and had better reliability and robust control characteristics.  For additional technical details of this project, please see Case Study 2.1

I look forward to the discussion and collaboration on this idea at the workshop, and I encourage all attendees to subscribe to the ARPA-E newsletter to stay in the loop on future happenings on this and other topic areas.

1. Garcia-Sanz, M. Robust Control Engineering: Practical QFT Solutions. (Boca Raton, Florida: CRC Press, 2017), 317-342.