The invention relates to a transformer noise suppression method, and belongs to the technical field of electric power supplying. The transformer noise suppression method comprises the steps that firstly, initial noise digital signals are received and converted to serve as input signals of the BP neural network; secondly, the input signals are processed to generate secondary digital signals; thirdly, the secondary digital signals are output to a loudspeaker, and secondary noise is generated; fourthly, remained noise digital signals obtained by overlapping the initial noise and the secondary noise are received, whether the number of times is continuously set not to be changed is judged, if yes, the secondary digital signals are output, and if not, BP neural network parameters are optimized and adjusted with the amplitude of the remained noise digital signals being minimum as the optimality principle, initial noise digital signals of the next time serve as new input signals of the BP neural network, and the second step is executed again. By means of the transformer noise suppression method, noise within a certain range of noise sources such as a transformer can be quickly suppressed, an ideal noise reduction effect can be achieved, and the problems that a traditional noise reduction method is high in cost, poor in effect and portability and the like are solved.