Construction project post-evaluation is the feedback link of the project decision management, and it can provide a scientific basis for construction projects in the future. This paper established the post-evaluation indicator system of power plant construction project according to its feature. Based on the improved back propagation neural network by using adaptive learning rate and momentum algorithm, combined with Delphi method and success degree method, the post-evaluation model was established. Through the empirical study on 300MW power plant construction projects, compared with the traditional back propagation neural network model, the result of improved BP neural network model is more accurate and effective.