Iron and steel enterprise multi-energy day-ahead scheduling optimization method considering multi-level indexes

Abstract

The invention provides an iron and steel enterprise multi-energy day-ahead scheduling optimization method considering multi-level indexes, and the method comprises the steps: determining multi-criterion evaluation indexes which comprise a total system index and a subsystem index; establishing an optimization model for the multi-energy system of the iron and steel enterprise according to the evaluation index of the total system and the energy pipe network condition; establishing an optimization model aiming at the gas subsystem according to the evaluation indexes of the subsystem and the actual condition of the subsystem; according to evaluation indexes and optimization models of different levels, defining an upper-layer optimization model of a total system level and a lower-layer optimization model of a subsystem level, and constructing a double-layer optimization model; and a double-layer cascade optimization solution strategy based on Actor-Critic reinforcement learning is adopted to carry out solution. According to the method, two layers of optimization models are considered in each step to search an optimal solution, an energy input allocation scheme in which subsystem evaluation indexes reach the standard and total system indexes are optimized is obtained, and a day-ahead scheduling optimization scheme considering multi-level indexes is provided for the iron and steel industry, so that an enterprise is helped to make a production plan.

Description

technical field [0001] The invention relates to the technical field of multi-energy scheduling optimization, in particular to a multi-energy day-ahead scheduling optimization method for iron and steel enterprises considering multi-level indicators. Background technique [0002] Iron and steel enterprises involve coal, natural gas, electricity, by-product gas and other energy sources, and the pipeline network structure is complex. An unreasonable day-ahead scheduling optimization scheme will lead to problems such as high production energy consumption and high pollution emissions. However, there are often contradictions between the overall optimization of the pipeline network and the local optimization of the energy subsystem. In order to ensure that the production plan is completed on schedule, each energy medium is often in an unbalanced state where supply exceeds demand. In order to reduce the energy loss of the entire system, iron and steel enterprises often try to recycl...

AI Technical Summary

Problems solved by technology

However, mathematical analytical methods represented by linear programming, nonlinear programming and dynamic programming are not suitable for multi-objective optimization problems with non-convex objective functions and complex nonlinear constraints; while conventional multi-objective intelligent optimization algorithms have strong coupling of variables The solution speed of complex system optimization problems is extremely slow, which cannot meet the needs of actual field applications

Images