Stack denoising autocoder-based probabilistic power flow online calculation method

Abstract

The invention discloses a stack denoising autocoder-based probabilistic power flow online calculation method. The method mainly comprises the following steps of 1) building an SDAE probabilistic powerflow model; 2) obtaining training samples of the SDAE probabilistic power flow model; 3) initializing the SDAE probabilistic power flow model; 4) training the SDAE probabilistic power flow model to obtain a trained SDAE probabilistic power flow model; 5) obtaining calculation samples; 6) inputting calculation sample data obtained in the step 5) to the trained SDAE probabilistic power flow model in the step 4 once, obtaining training targets, thereby judging power flow solvability of all the training samples, and calculating power flow values of the solvable samples; and 7) performing statistics on probabilistic power flow indexes. The method can be widely applied to the probabilistic power flow online calculation of a power system, and is especially suitable for the situation of power system uncertainty enhancement due to high-ratio access of new energy.

Description

technical field [0001] The invention relates to the field of electric power system and automation thereof, in particular to an online calculation method of probability power flow based on a stacked noise-reduction autoencoder. Background technique [0002] Power systems essentially operate in uncertain environments. Probabilistic power flow can take into account the influence of uncertain factors, obtain the probability characteristics of system state variables, and use it in power system planning and operation. In recent years, due to the increasing penetration of renewable energy such as photovoltaics and wind power, the uncertainty of the power system has surged. In order to meet the requirements of power system operation and scheduling, the demand for online probabilistic power flow calculation is becoming more and more urgent. [0003] At present, the probabilistic power flow solution methods mainly include analytical method and simulation method. Analytical methods ...

AI Technical Summary

Problems solved by technology

Most of the improved iterative algorithms are based on Newton's method, such as fast decoupling method, quasi-Newton method, etc., which speed up the power flow solution to a certain extent, but iterative calculation is still required, so it is difficult to be used for online analysis

Improved non-iterative algorithms, such as DC power flow algorithms, traditional neural network algorithms, etc., have demonstrated the ability of online calculations, but they have shortcomings such as low power flow calculation accuracy

Images