Research On Distributed Computing And Convex Relaxation Technology For Optimal Energy Flow Problem Of Integrated Energy System

In order to achieve the “carbon peak” and “carbon neutral” strategic goals,it is necessary to promote low-carbon energy transition and energy revolution.The integrated energy system is an important way for power system construction and energy reform.It improves energy efficiency and the proportion of renewable energy consumption by taking advantage of the complementary and alternative characteristics of electricity and gas,but it correspondingly increases the difficulty of system scheduling and operation.On the one hand,each system scheduling subject has requirements for information protection and autonomous decision-making,which leads to limitations in the traditional centralized scheduling method.On the other hand,there are non-convex equations in the system model,and the non-convex characteristics make it difficult to solve system scheduling and distributed optimization problems.To this end,this article is based on distributed computing and convex relaxation technology,starting from the distributed optimal power flow problem of the traditional multi-regional interconnected power system,and then studies the distributed optimization,convex optimization and energy flow analysis of the optimal energy flow problem of the integrated energy system.The main research content of this paper is summarized as follows.First,the application of distributed computing and convex relaxation technology in the optimal energy flow problem is studied.This part of the research content can be used to deal with the shortcomings of the centralized dispatching method and the non-convex characteristics of the model in the optimal energy flow problem.It can lay the foundation for the research on the optimal power flow of the multi-region interconnected power system and the optimal energy flow of the integrated energy system in the later chapters of this article.Second,based on the above research methods,for the optimal power flow problem of the multi-area interconnected power system,a two-layer solution algorithm for the multi-objective distributed optimal power flow of the power system based on convex relaxation iterative method is proposed.The upper-layer optimization is a sequential optimization algorithm,using convex relaxation iterative method to deal with non-convex constraints;the lower-layer optimization is a distributed algorithm,which can be used for distributed scheduling calculations.The proposed algorithm solves the problems caused by the centralized dispatching method and the non-convex characteristics of the model,and is suitable for solving the distributed optimal power flow problem of any topology grid.Third,based on the above research methods,a convex relaxation iterative solution algorithm for the distributed optimal energy flow of the integrated electricity-gas system considering the dynamic characteristic of line pack is proposed in the context of the integrated energy system.In terms of model,a natural gas network model description method without integer variables considering the dynamic characteristic of line pack is studied,which can improve the calculation accuracy and efficiency of the model.In terms of algorithm,the solution process transforms the non-convex optimal energy flow problem into an iteratively solved convex programming problem through the sequential optimization algorithm,and then uses the distributed algorithm at each step of the sequential optimization to perform distributed collaborative optimization of the entire system to achieve “decentralized autonomy,regional coordination”.