Realizing and continuously guaranteeing operational performance indicators (OPIs), is a challenging task for multi-physics (electro, thermal and mechanical) systems and complex mechatronic systems, as these are subject to multiple objectives (fast vs qualitative production), varying context (changing loads and equipment) and multiple interactions (heating depending on air inlet flow, heat sources nearby).  To cope with these challenges, corrections and adaptations in the control settings are typically performed, for instance through regular adjustment of a set of individual controllers by a supervisory controller, or through retuning the control system by predefined functions designed from historic data and numerical models.

Instead, the ADCEE tools are developed to assist the controller by increasing the awareness of the system of its own behaviour and its impact on the (optimal) performance. For this, the ADCEE tools are based on the principles of dual control, a control strategy that aims on one hand to exploit system knowledge, for optimal performance, and on the other hand to explore the system for rich system identification. The ADCEE tools attempt to identify system parameters and states relevant in an online systematic way for a more effective adaptation of the control settings. However, in current dual control, this balance of the control efforts between exploration and exploitation is often fixed, chosen offline, performed for the individual controllers, and is not developed to deal with variable objectives, different system operating scenarios or varying loads.

The project ADCEE takes the next step in this dual control by defining and developing an adaptive dual control, perfectly balancing decisions on the most suitable control set points for both exploration and exploitation. It creates the extra functionality within the controller for a control-oriented system awareness, aiming at identifying and upholding the highest performance through effective adaptation of control settings.