The article discusses how to predict the cost of power engineering projects more accurately using historical data. By using a type of artificial intelligence called a BP neural network, researchers created a model to quickly analyze engineering costs. The study showed that this model is feasible and effective for predicting the costs of new power projects.