Abstract: A day ahead electricity price estimation method is proposed by combining the convolutional neural networks (CNN) with evolutionary algorithm. Based on the results of gene mapping, competitive individuals were selected, mutated and crossed. The optimal CNN structure was designed according to the mapping relationship between gene bits and CNN parameters and hyper parameters. A power market was taken as an example to verify the effectiveness of the proposed method. Experimental results show that the proposed algorithm has higher prediction accuracy and lower error rate than other prediction methods.