Part of the investment data of a power project.

In view of the shortcomings of power engineering cost in precision and dynamic in big data environments, this paper proposes building information modelling (BIM) and spatiotemporal modelling-based dynamic graph convolutional neural networks (DynGCN). This study uses the characteristics of BIM technology to carry out the cost management of the whole life cycle of power engineering, and realizes the dynamic control of the cost. In addition, the DynGCN method is used to predict the cost of each engineering link, so as to optimize the construction scheme of the whole project. The results show that the whole life cycle data management supported by BIM technology improves the real-time monitoring and adjustment ability of the cost; the DynGCN method can greatly improve the accuracy of the cost prediction, and the prediction accuracy is 96%, which is closest to the real value of the cost.