Gas supply and demand dynamic prediction system for steel enterprises and method thereof

Abstract

The invention provides a gas supply and demand dynamic prediction system for steel enterprises and a method thereof, belonging to the technical field of gas prediction in steel industry. The system comprises a data acquisition subsystem, a data processing subsystem, a data modeling subsystem, a model verification subsystem, a model application subsystem and a computer network connecting each subsystem. The gas supply and demand dynamic prediction method comprises the following steps: abstracting generation and consumption characteristics of gas; qualitatively analyzing the generation and consumption characteristics of the gas; dynamically predicting and modeling in a sectional short-term method; and dynamically predicting the supply and demand of the gas. The system and the method of the invention have the advantages of proposing the sectional short-term dynamic prediction and modeling method which is based on historic statistic data, combines the production state information and comprehensively applies real-time data, the production state data and production process technical data. The invention provides decision-making support for distributing staff in short-term gas distribution, and simultaneously provides basic data for the implementation of various intelligent distribution schemes.

Description

technical field [0001] The invention belongs to the technical field of gas forecasting in the iron and steel industry, and in particular provides a dynamic forecasting system and method for gas supply and demand in iron and steel enterprises, which are used for short-term gas forecasting in iron and steel enterprises. Background technique [0002] At present, the research on gas forecasting in the iron and steel industry is a static forecast made in a relatively long period of time, without considering the production plan and equipment maintenance plan, that is, the forecast is made according to the normal development, without considering the impact of external events, Effective real-time dynamic forecasting cannot be realized, so how to conduct online dynamic forecasting based on dynamic working condition information and provide support for dynamic optimal scheduling is a very important research field. [0003] The current systems and methods for dynamic forecasting of gas ...

AI Technical Summary

Problems solved by technology

[0002] At present, the research on gas forecasting in the iron and steel industry is a static forecast made in a relatively long period of time, without considering the production plan and equipment maintenance plan, that is, the forecast is made according to the normal development, without considering the impact of external events, It is impossible to realize effective real-time dynamic prediction. How to conduct online dynamic prediction based on dynamic working condition information and provide support for dynamic optimal scheduling is a very important research field

[0003] The current systems and methods for dynamic forecasting of gas supply and demand, such as the invention patent "An Integrated Online Energy Forecasting System and Method for Iron and Steel Enterprises", use a variety of energy forecasting algorithms and a combined model for energy forecasting, which overcomes the need for single or One or two energy forecasting methods to predict the limitations of energy demand, improve the accuracy and reliability of energy forecasting Metallurgical Automation Research and Design Institute ZL200610113685.6, but the system and method provided by the invention patent do not consider production planning, maintenance planning, and production conditions information, that is, to predict according to normal development without considering the influence of external events, so the prediction results have a certain deviation from the actual situation. The invention patent "an online energy forecasting system and method based on the product ARIMA model" proposes a An online energy prediction technology based on the ARIMA method, which is suitable for various energy data types including stationary, non-stationary, and seasonal fluctuations. Metallurgical Automation Research and Design Institute 200810226961.9, but the model provided by this invention patent is too simple to adapt to the complex and changeable reality of the iron and steel industry At the same time, the consideration of production plan, maintenance plan, and production condition information is limited, and the actual production cannot be well simulated, so the model has certain limitations. The document "Statistical Research on Energy Forecasting Methods of Enterprises" uses the statistical model of energy consumption Energy Research and Utilization of Energy Forecasting 1993, this document uses statistical regression model to predict energy demand, the regression model method has the following advantages: simple and practical, it can not only predict energy demand, but also can be used in various factors that affect energy demand , using correlation tests to determine the most important influencing factors, thereby simplifying the model and highlighting the main contradiction

However, using the regression model has the following obvious disadvantages: (1) When the regression equation is used for estimation and prediction, the dependent variable can only be estimated from the independent variable, and the independent variable is not allowed to be estimated from the dependent variable

(3) The regression equation should only be used for interpolation calculations, not for extrapolation predictions, especially for remote extrapolation predictions

At the same time, there are obviously the following deficiencies; ①G(1,1) model is an exponential growth model, when a short time series is used to predict a long time series, it will produce large errors or values ​​that do not conform to the actual situation , ②For the original data containing negative value items, if the data generated after multiple accumulations cannot obtain non-negative incremental data, theoretically speaking, the modeling of such data should be abandoned, ③When the GM(1,1) model When the accuracy cannot meet the requirements, it is necessary to build a GM(1,1) model on the residual to correct the original model to improve the accuracy. In many practical problems, the residual contains both positive and negative items, and the accumulated data generated It is definitely not non-negative increasing, so the original model cannot be modified

Although the neural network model has a high nonlinear mapping ability and can approximate nonlinear functions with arbitrary precision, there are still some problems in actual calculation: ① The calculation process of backpropagation has a slow convergence speed, and generally requires hundreds of Thousands of iterative calculations; ②There is a minimum value of the energy function; ③The selection of the number of hidden neurons and connection weights often depends on experience; ④The convergence of the network is related to the structure of the network, etc.

Images