Oxygen Load Prediction in Steel Plants via Neural Network
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current oxygen load prediction methods in iron and steel enterprises rely solely on historical data, failing to accurately predict changes due to sudden shifts in production plans, leading to inefficient oxygen utilization and production disruptions.
Innovation Solution
A method that combines production planning data with historical performance to establish a neural network model for predicting oxygen consumption, using key influencing factors to forecast oxygen load in time granularity, ensuring accurate prediction and optimization of oxygen usage across the enterprise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If data-driven methods (neural network, SVM, fuzzy system) are used to predict oxygen load based on historical data, then prediction can be performed in time granularity, but the prediction accuracy deteriorates when production plan changes suddenly
Solution Approach 1:
The patent applies preliminary action by extracting key features from the production plan (such as converter blowing schedule, oxygen consumption requirements) before the actual oxygen consumption occurs. These pre-extracted features are used to guide the prediction process, allowing the system to anticipate oxygen load changes based on planned production activities rather than relying solely on historical patterns. This enables accurate prediction even when production plans change suddenly.
2Ease of manufacture
If oxygen consumption is predicted using only historical data without production plan information, then the prediction method is simple, but the prediction cannot respond to sudden changes in production plan
Solution Approach 1:
The patent introduces production plan information as an intermediary element that bridges historical data and future oxygen consumption prediction. The production plan serves as a mediator that translates production scheduling decisions into predicted oxygen load patterns. By incorporating this intermediary, the system maintains the simplicity of data-driven methods while significantly improving reliability under plan changes, as the production plan directly informs the prediction model about upcoming oxygen requirements.
3Reliability
If oxygen supply is increased to ensure stable system pressure during intensive converter blowing, then system pressure stability is maintained, but oxygen waste increases due to oversupply
Solution Approach 1:
The patent implements feedback by using the predicted oxygen consumption information to dynamically adjust oxygen supply in real-time. The prediction model continuously monitors planned production activities and feeds this information back to the oxygen supply control system. This feedback mechanism enables the system to match oxygen supply precisely with actual consumption needs, maintaining system pressure stability while eliminating oxygen waste from oversupply situations.
Solution Approach 2:
The patent applies dynamics by transitioning from static oxygen supply strategies to dynamic, prediction-driven supply adjustment. The system continuously updates oxygen supply levels based on real-time production plan changes and predicted consumption patterns. This dynamic approach allows the system to adapt oxygen supply to actual needs, maintaining pressure stability during intensive blowing while reducing waste during lower-demand periods.
4Reliability
If air compressor units operate at fixed load to ensure stable operation, then operational stability is maintained, but the system cannot meet sharp increases in oxygen demand
Solution Approach 1:
The patent applies preliminary action by predicting oxygen demand spikes in advance based on the production plan. When the model detects upcoming intensive converter blowing operations, it generates early warnings that allow air compressor units to be pre-positioned or pre-started. This preliminary preparation ensures that when sharp increases in oxygen demand occur, the system can respond immediately without compromising operational stability, as the compressors are already ready to deliver the required oxygen flow.
Data Source
AI summary
The present disclosure discloses a method for predicting oxygen load in iron and steel enterprises based on production plan, which relates to influencing factor extraction, neural network modeling and similar sequence matching technologies. The method uses the actual industrial operation data to first extract the relevant data such as the production plan and production performance of converter steel-making, analyze the influencing factors, and extract the main influencing variables of oxygen consumption. Then, the neural network prediction model of oxygen consumption of a single converter is established, the mean square error is taken as the evaluation index, and the predicting result of time granularity of a converter in the blowing stage is given. Finally, in combination with the information of smelting time and smelting duration of each device in the converter production plan, the prediction value of oxygen load in a planned time period is given.