A Practical Multiobjective Learning Framework for Optimal Hardware–Software Co-Design of Control-on-a-Chip Systems

The digital age has made embedded control a key component to user-oriented, portable, and the Internet of Things (IoT) devices. In addition, with emergent complex systems, there is a need for advanced optimization-based control strategies such as model predictive control (MPC). However, the unified implementation of these advanced strategies on hardware remains a challenge. Designing complex control policies for embedded systems is inherently an interwoven process between the algorithmic design and hardware implementation, which will require a hardware-software co-design perspective. We propose an end-to-end framework for the automated design and tuning of arbitrary control policies on arbitrary hardware. The proposed framework relies on deep learning as a universal control policy representation and multiobjective Bayesian optimization (MOBO) to facilitate iterative systematic controller design. The large representation power of deep learning and its ability to decouple hardware and software design are a central component to determining feasible control-on-a-chip (CoC) policies. Then, Bayesian optimization (BO) provides a flexible sequential decision-making framework where practical considerations, such as multiobjective optimization (MOO) concepts and categorical decisions, can be incorporated to efficiently design embedded control policies that are directly implemented on hardware. We demonstrate the proposed framework via closed-loop simulations and real-time experiments on an atmospheric pressure plasma jet (APPJ) for plasma processing of biomaterials.