Integrated electrical and natural-gas energy system probability energy flow calculation method based on stacked denoising autoencoder

Abstract

The invention discloses an integrated electrical and natural-gas energy system probability energy flow calculation method based on a stacked denoising autoencoder. The method comprises the following steps that: firstly, by virtue of the deep stack structure and the coding and decoding process of an SDAE (Stacked Denoising Autoencoder), establishing an energy flow model based on the SDAE to effectively mine the high-order characteristics of a nonlinear energy flow equation; on the basis, combining with numerical value characteristics input and output by the energy flow to put forward a trainingmethod based on a ReLU activation function, deviation standardization normalization and a small-batch gradient descent method to improve training accuracy and speed; and then, combining with an MCS (Modulation and Coding Scheme) method to sample a sample to be solved, and using a trained SDAE energy flow model to directly map the energy flow values of all sampled samples so as to realize the high-accuracy online calculation of the probability energy flow on a premise that hardware cost is not increased.

Description

technical field [0001] The invention belongs to the field of comprehensive energy systems. Background technique [0002] In order to promote new energy consumption, improve energy utilization efficiency, and achieve energy conservation and emission reduction, vigorously developing Integrated Electrical and Natural-Gas Energy Systems (IENES) has become an important trend in the development of the energy field. IENES inherently operates in an uncertain environment. Electric-gas probabilistic energy flow analysis can take into account various uncertainties and coupling characteristics of electric power and natural gas systems, and obtain the probability characteristics of system state quantities, which is the basis for scientific analysis of IENES. In recent years, due to the increasing penetration of renewable energy sources such as photovoltaics and wind power, system uncertainty has surged. In order to meet the requirements of IENES operation scheduling, the demand for onl...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to solve the problem that probabilistic energy flow needs to solve a large number of highly nonlinear equations, resulting in high calculation cost and long solution time, and it is difficult to popularize and apply it in actual engineering. An online method of probabilistic energy flow based on SDAE and combined with MCS method is proposed , to provide technical support for the online application of probabilistic energy flow

Images