Coupled calibration-control method for complex systems

We formulate the calibration-control dilemma which occurs in most interesting out-of-equilibrium dynamical systems in the natural and social sciences. Despite a lack of knowledge of the main control parameters, standard control methodology assumes that calibration can determine them. They are, however, obfuscated by the existence of control, leading to large undesired intermittent bursts, crises, and instabilities. By merging parameter calibration and control in a data assimilation framework, our joint calibration-control method nearly matches the performance achieved with complete knowledge of the system parameters, as demonstrated by both analytical and numerical simulations. This paves the way toward more versatile control of complex systems.