A method for optimizing the layout of magnetic sensors in tokamaks based on improved minimal-redundancy-maximal-relevance criterion

The magnetic diagnostic system provides critical input signals for plasma feedback control, which is vital for the safe and stable operation of tokamak devices. Considering system cost and engineering constraints related to installation, optimizing the placement of magnetic sensors to minimize the sensor count is essential. In this study, a sensor layout optimization algorithm based on an improved minimal-redundancy-maximal-relevance criterion is proposed. This improved criterion takes into account both the distance and measurement direction differences between sensors in the redundancy assessment, addressing the issue of dense sensor placement in localized regions, often observed in existing studies. In addition, binary search is employed to identify the minimum number of sensors required, significantly speeding up the optimization process. This algorithm is applied to the design of magnetic sensor layout for a next-generation fusion energy experimental device. The optimized layout consists of only 13 pick-up coils and 16 flux loops. In contrast, a widely used uniform layout requires approximately 50 sensors to achieve the same reconstruction accuracy, demonstrating the effectiveness and practicality of the proposed optimization algorithm.