模式识别非线性映照技术在配矿优化中的应用

对钢铁生产中多目标调优的不良结构问题，试采用交互模式识别的非线性映照信息处理技术，寻找优化目标的类别属性与影响因素值间的对应关系，从而解析优类样本的共性。此方法应用于高炉多项经济、技术指标优化的探索，由计算结果知：入炉品位是影响高炉经济、技术指标的最重要因素，而提高烧结矿品位是提高武钢入炉品位的最佳方法。应用于生产后，已收到显著效果。

Non-linear mapping information processing technology of interactive pattern recognition and its application of optimizing ore mixture proportion

Trying to solve the inappropriate structural problem for multiple objective optimizations in iron and steel production, non-linear mapping information processing technology through interactive pattern recognition is used. This technology finds out according relationship between the characterizations of optimized objectives and the values of affecting factors, so the commonality in excellent sample can be figure out. The method has been applied in researches for several economical and technical optimizations of furnace, so we can come to a conclusion after calculating that iron grade is the most important affecting factor for economical and technical aims of furnace, and promoting the grade of sintering ore is the best way to Optimizing Ore Mixture Proportion. The conclusion turns to be effective when using in the practice.