【摘要】 受限于炼钢环境，直接测量钢水温度难以实现。RH精炼炉温度控制影响整个钢铁冶炼生产的效率和质量。为了尽可能准确地预测钢液的终点温度，文章提出了一种基于食肉植物算法(CPA)优化的支持向量回归(SVR)炉外终点温度预测模型。以实际生产中的炼钢记录作为数据支撑，进行仿真实验。结果表明，文章提出的CPA-SVR模型预测误差|Δt|≤5℃条件下的命中率为93.442 6%,优于多元线性回归模型、BP模型和SVR模型，对实际生产中预测钢水的终点温度具有一定的指导意义。更多还原

【Abstract】 Limited by the steelmaking environment, it is difficult to measure the molten steel temperature directly. Temperature control affects the efficiency and quality of the whole iron and steel smelting production.In order to predict the end-point temperature of molten steel as accurately as possible, a support vector regression(SVR) end-point temperature prediction model based on carnivorous plant algorithm(CPA) optimization is proposed. Taking the actual steelmaking records as the data support, the simulation experiment is carried out. The results show that the hit rate of the CPA-SVR model proposed in the paper under the condition of prediction error |Δt|≤5 ℃ is 93.442 6%, which is better than multiple linear regression model, BP model and SVR model. It has certain guiding significance for predicting the end temperature of molten steel in actual production.更多还原