FUsion Synthesis Engine: a next-generation framework for integrated design of fusion pilot plants (FPPs)

The FUsion Synthesis Engine (FUSE) framework represents a breakthrough in fusion plant design, integrating first-principle, machine-learning, and reduced models into a comprehensive simulation tool for physics, engineering, control, and costing. Traditional fusion pilot plant (FPP) design methods utilize low-fidelity simplified analytical models, identifying an operational point for further high-fidelity iterative analysis by physicists and engineers. FUSE surpasses these traditional methods by providing self-consistent physics/engineering solutions as a medium-fidelity starting point. High-fidelity physics models and engineering analysis-driven design changes can be directly fed back into FUSE, enabling an updated integrated design. This approach minimizes the need for time-consuming expert evaluations and significantly reduces human involvement in the design iteration loop. FUSE, designed with a focus on computational capabilities, modularity, flexibility, and extensibility, embraces modern software practices. It leverages the ITER IMAS data structure for effective data exchange, thereby facilitating validation and verification of simulated conceptual designs against experiments and existing predictive tools. Employed in a wide range of studies, from component design to plant-wide optimization, FUSE accommodates uncertainty quantification, thus streamlining the fusion plant design process and aiding in risk-informed decision making.