To address the issues of real-time measurements shortage after large-scale integration of high-proportion distributed generation (DG) and the limited estimation accuracy of traditional forecasting-aided state estimation (FASE) methods

a FASE method of active distribution network based on improved Crossformer pseudo-measurements modeling is proposed. Firstly

highly correlated input features are selected on the basis of the maximal information coefficient (MIC) to enhance the accuracy of the prediction model. Then

the total variation regularized (TV) technique is employed to optimize the robust principal component analysis (RPCA) for constructing the TRPCA layer

which is embedded into Crossformer to address its incapability in effectively handling non-Gaussian noise. Lastly

the improved prediction model is utilized for ultra-short-term load forecasting

and pseudo-measurements of bus states are obtained through power flow calculation to supplement missing data in case of insufficient measurements. The extended Kalman filter (EKF) is employed for state estimation. Simulation tests on the IEEE 33-bus and IEEE 118-bus test systems verify the advantages of the proposed method in terms of estimation accuracy and robustness

providing a reference for active distribution network FASE.