首页|Reports from General Atomics Add New Data to Findings in Machine Learning (Augme nting Machine Learning of Grad-shafranov Equilibrium Reconstruction With Green’s Functions)
Reports from General Atomics Add New Data to Findings in Machine Learning (Augme nting Machine Learning of Grad-shafranov Equilibrium Reconstruction With Green’s Functions)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News – Current study results on Machine Learn ing have been published. According to news reporting originating from San Diego, California, by NewsRx correspondents, research stated, “This work presents a me thod for predicting plasma equilibria in tokamak fusion experiments and reactors . The approach involves representing the plasma current as a linear combination of basis functions using principal component analysis of plasma toroidal current densities (J(t)) from the EFIT-AI equilibrium database.” Funders for this research include U.S. Department of Energy10.13039/100000015, U nited States Department of Energy (DOE). Our news editors obtained a quote from the research from General Atomics, “Then utilizing EFIT’s Green’s function tables, basis functions are created for the po loidal flux ( psi) and diagnostics generated from the toroidal current (J(t)). S imilar to the idea of a physics-informed neural network (NN), this physically en forces consistency between psi, J(t), and the synthetic diagnostics. First, the predictive capability of a least squares technique to minimize the error on the synthetic diagnostics is employed. The results show that the method achieves hig h accuracy in predicting psi and moderate accuracy in predicting J(t) with media n R-2 = 0.9993 and R-2 = 0.978, respectively. A comprehensive NN using a network architecture search is also employed to predict the coefficients of the basis f unctions. The NN demonstrates significantly better performance compared to the l east squares method with median R-2 = 0.9997 and 0.9916 for J(t) and psi, respec tively. The robustness of the method is evaluated by handling missing or incorre ct data through the least squares filling of missing data, which shows that the NN prediction remains strong even with a reduced number of diagnostics.”
San DiegoCaliforniaUnited StatesNo rth and Central AmericaCyborgsEmerging TechnologiesMachine LearningGener al Atomics
NETL
NSTL
