New energy power state estimation method and system based on deep learning

Abstract

The deep learning-based new energy power system state estimation method and system comprises the following steps: step 1, a new energy power system state estimation method and system based on an improved DNN are proposed; step 2, the RTU data of step 1 is predicted based on BILSTM, then the Spline interpolation is used to generate pseudo-measurement, and then the PMU measurement of the node measurement is fused; step 3, based on the fusion data set obtained in step 2, a data identification technology based on the joint space-time FC mechanism and BILSTM of compatible measurement is proposed; step 4, based on the data identification technology proposed in step 3; through deep identification of the data, the bad data is removed, and the fluctuation data is retained, so that the original data set is obtained; step 5, a new energy power system state estimation model is established according to the original data set obtained in step 4. The present application improves the accuracy and speed of the new energy power system state estimation on the basis of retaining the original characteristics of the new energy measurement data, and is more suitable for the needs of the development of new power systems.

Images