Abstract: In order to improve the energy utilization efficiency and injection power of the discharge plasma,a BP neural network optimized by genetic algorithm with the inputs of discharge voltage,discharge frequency,discharge length and gas flow rate was established to predict the power factor and discharge power. With the help of model prediction data,the influence laws of the above four factors on power factor and discharge power were explored to guide the practical application and optimal design of dielectric barrier discharge. According to the test set data,the average error level of the discharge power was 3%,the average error level of the power factor was 2.73%,indicating that the model had high prediction accuracy and fast convergence speed.