The Back Propagation Neural Network (BPNN) has been applied to the density inversion problem of the POlarimeter INTerferometer (POINT) system on the EAST tokamak. Using the BPNN, the electron density profile can be directly reconstructed from the line-integrated density measurement provided by the POINT system. The accuracy and reliability of this approach have been investigated through tests on experimental data. Compared to the traditional Park-matrix method, the BPNN-based model demonstrates significantly faster performance and greater robustness against system noise, making it suitable for real-time control of the density profile. Additionally, the influence of various measurement channels on the inverted density profile has been thoroughly analyzed, offering a quantitative approach to optimizing interferometer design for future machines.
This paper presents a neural network-based method to reconstruct the electron density profile from interferometer measurements on the EAST tokamak. The approach is faster and more robust to noise compared to traditional methods, making it suitable for real-time control of the density profile.