Prediction of soft X-ray laser gain value generated from laser plasmas by using a multilayer perceptron neural network

The hot plasma generated from the high-power pulse of the laser on the target surface is considered such as a suitable source for the amplification of soft X-ray lasers. The gain coefficient and output intensity of these types of lasers depend on the characteristics of the input pulse and the target material. Optimizing emission for practical purposes requires extensive numerical simulations based on hydrodynamic equations and equation of state, which is computationally difficult and complex. In this research, a solution to this problem is proposed by using a model based on machine learning to predict the gain value of soft X-ray laser with a multilayer perceptron neural network, which there is no need to directly numerically solve an emission model. Finally, it can be seen that the neural network can predict the gain coefficient of the soft X-ray laser relatively well using only the features of pumped pulse. The use of neural networks in soft X-ray laser emission produced from laser plasmas is a new approach that needs further investigations according to the extent of simulation and optimization algorithms.