With the gradual expansion of the scale of the charging station, the influence of the harmonic current on the distribution network becomes particularly serious when it is put into use. In the practical application, the traditional national standard harmonic superposition algorithm has a deviation from the actual harmonics because of the influence of various factors in the charging station. To solve this problem, the variational Bayesian posterior parameter estimation method is applied to harmonic current superposition algorithm, this method by measuring lower bound space continuously maximum marginal likelihood function in the state space and the amount of harmonic phase, iterative updates of variational phase parameters, until the approximate distribution of the true posterior approximation parameter distribution, from the estimated parameters to achieve harmonic phase superposition. The experimental simulation shows that the variational Bayesian estimation method is more effective and accurate for the superposition of the phase parameters which obey Gauss distribution than the traditional harmonic superposition algorithm, and provides an effective basis for the precise harmonic suppression.