Abstract: Polynomial chaotic expansion methods have shown considerable potential for analyzing the uncertainty of stochastic plasma electromagnetic propagation. However, because the amount of computation required to construct a polynomial chaotic model increases exponentially as the dimension of the uncertainty input variable increases, numerical simulations are time-consuming, resulting in the problem of "dimensional disaster". Thus, based on nonintrusive polynomial chaos(NIPC) method, the hybrid Monte Carlo/polynomial chaotic expansion(MC/PCE) method is used to study the effect of the multi-layer plasma flat-panel electron density uncertainty on transmission coefficient, which demonstrates the practicability of the proposed approach. By comparing it to the MC method, which can greatly accelerate the simulation speed, it is improved that the method effectively alleviates the rapid increase of the number of polynomial expansion terms with the increase of the dimension of random variables, and greatly reduces the number of simulations compared with Monte Carlo method, which is conducive to the multi-parameter uncertainty analysis of electromagnetic models.