Deep learning approaches to recover the plasma current density profile from the safety factor based on Grad–Shafranov solutions across multiple tokamaks

Many magnetohydrodynamic stability analyses require generation of a set of equilibria with a fixed safety factor q-profile while varying other plasma parameters.A neural network(NN)-based approach is investigated that facilitates such a process.Both multilayer perceptron(MLP)-based NN and convolutional neural network(CNN)models are trained to map the q-profile to the plasma current density J-profile,and vice versa,while satisfying the Grad–Shafranov radial force balance constraint.When the initial target models are trained,using a database of semianalytically constructed numerical equilibria,an initial CNN with one convolutional layer is found to perform better than an initial MLP model.In particular,a trained initial CNN model can also predict the q-or J-profile for experimental tokamak equilibria.The performance of both initial target models is further improved by fine-tuning the training database,i.e.by adding realistic experimental equilibria with Gaussian noise.The fine-tuned target models,referred to as fine-tuned MLP and fine-tuned CNN,well reproduce the target q-or J-profile across multiple tokamak devices.As an important application,these NN-based equilibrium profile convertors can be utilized to provide a good initial guess for iterative equilibrium solvers,where the desired input quantity is the safety factor instead of the plasma current density.

Loss of energetic particles due to feedback control of resistive wall mode in HL-3

Effects of three-dimensional(3D)magnetic field perturbations due to feedback control of an unstable n=1(n is toroidal mode number)resistive wall mode(RWM)on the energetic particle(EP)losses are systematically investigated for the HL-3 tokamak.The MARS-F(Liu et al 2000 Phys.Plasmas 73681)code,facilitated by the test particle guiding center tracing module REORBIT,is utilized for the study.The RWM is found to generally produce no EP loss for cocurrent particles in HL-3.Assuming the same perturbation level at the sensor location for the close-loop system,feedback produces nearly the same loss of counter-current EPs compared to the open-loop case.Assuming however that the sensor signal is ten times smaller in the close-loop system than the open-loop counter part(reflecting the fact that the RWM is more stable with feedback),the counter-current EP loss is found significantly reduced in the former.Most of EP losses occur only for particles launched close to the plasma edge,while particles launched further away from the plasma boundary experience much less loss.The strike points of lost EPs on the HL-3 limiting surface become more scattered for particles launched closer to the plasma boundary.Taking into account the full gyro-orbit of particles while approaching the limiting surface,REORBIT finds slightly enhanced loss fraction.

Effect of ideal internal MHD instabilities on NBI fast ion redistribution in ITER 15 MA scenario

Transport of fast ions is a crucial issue during the operation of ITER.Redistribution of neutral beam injection(NBI)fast ions by the ideal internal magnetohydrodynamic(MHD)instabilities in ITER is studied utilizing the guiding-center code ORBIT(White R B and Chance M S 1984Phys.Fluids 272455).Effects of the perturbation amplitude A of the internal kink,the perturbation frequency f of the fishbone instability,and the toroidal mode number n of the internal kink are investigated,respectively,in this work.The n=1 internal kink mode can cause NBI fast ions transporting in real space from regions of 0<s≤0.32 to 0.32<s≤0.53,where s labels the normalized plasma radial coordinate.The transport of fast ions is greater as the perturbation amplitude increases.The maximum relative change of the number of fast ions approaches 5%when the perturbation amplitude rises to 500 G.A strong transport is generated between the regions of 0<s≤0.05 and 0.05<s≤0.12 in the presence of the fishbone instability.Higher frequency results in greater transport,and the number of fast ions in 0<s≤0.05 is reduced by 30%at the fishbone frequency of 100 k Hz.Perturbations with higher n will lead to the excursion of fast ion transport regions outward along the radial direction.The loss of fast ions,however,is not affected by the internal MHD perturbation.Strong transport from 0<s≤0.05 to 0.05<s≤0.12 does not influence the plasma heating power of ITER,since the NBI fast ions are still located in the plasma core.On the other hand,the influence of fast ion transport from 0<s≤0.32 to 0.32<s≤0.53 needs further study.

Synergistic Influences of Kinetic Effects from Thermal Particles and Fast Ions on Internal Kink Mode

The kinetic effects of thermal particles and fast ions on internal kink(IK)mode are numerically investigated by the MHD-kinetic hybrid code MARS-K.It is shown that either thermal particles or fast ions have stabilizing influence on IK.However,the former can not fully stabilize IK,and the later can suppress the IK.In addition,the synergistic effect from thermal particles and fast ions induces more stronger damping on IK.The kinetic effects from particles significantly raise the critical value of poloidal beta(βpcrit)for driving IK in the toroidal plasma.This implies a method of controlling IK or sawtooth in the high-βp discharge scenario of tokamak.It is noted that,at the q=1 rational surface,mode structure becomes more sharp due to the self-consistent modification by particles’kinetic effect.

Biogeographic distribution patterns of algal community in different urban lakes in China: Insights into the dynamics and co-existence

Urban lake ecosystems are significant for social development,but currently we know little about the geographical distribution of algal community in urban lakes at a large-scale.In this study,we investigated the algal community structure in different areas of urban lakes in China and evaluated the influence of water quality parameters and geographical location on the algal community.The results showed that obvious differences in water quality and algal communities were observed among urban lakes in different geographical areas.Chlorophyta was the dominant phylum,followed by cyanobacteria in all areas.The network analysis indicated that algal community composition in urban lakes of the western and southern area showed more variations than the eastern and northern areas,respectively.Redundancy analysis and structural equation model revealed that nutrients and p H were dominant environmental factors that affected the algal community,and they showed higher influence than that of iron,manganese and COD Mn concentration.Importantly,algal community and density exhibited longitude and latitude relationship.In general,these results provided an ecological insight into large-scale geographical distributions of algal community in urban lakes,thereby having potential applications for management of the lakes.