焦炉火道温度稳定性优化控制仿真

Optimization Control of Temperature Stability of Coke Oven

焦炉立火道温度的稳定性直接关系到炼焦生产的产量和质量。火道温度的稳定性将直接导致焦炭质量的下降、生产成本的提高。而现在焦化厂焦炉火道温度不能实时在线检测且设定值完全依靠人工经验给定,无法根据现场炉况变化做出实时调整致使焦炭生产不稳定,能耗大,生产成本高。针对这一现状提出基于软测量与设定值优化的火道温度闭环控制方案。设计中采用基于案例推理对焦炉火道标准温度进行寻优。结合海量现场数据和专家系统建立优化模型得到能耗小、产高的目标火道标准温度值。采用最小二乘支持向量机模型实现火道温度软测量。用小波神经网络进行焦炉立火道温度的预测。并通过仿真研究验证了方案的有效性。仿真结果表明该方案可以找出最佳标准温度设定值使煤气消耗量减少,能耗降低。可以为实际生产提供良好的指导。

The stability of the coke oven temperature directly relate to the production and quality of coking produc- tion, which will directly lead to the decline in the quality of coke and increase production cost. At present, coke oven temperature can not be measured online in real-time in the coking plant, and the coke oven temperature setting is complete dependent on human experience. It can not be adjusted in real-time according to the practical coke oven condition change. This makes coke production unsteadiness, high energy consumption and production costs. Aiming at this situation, we proposed a closed-loop temperature control scheme that bases on soft-sensing and optimization to set the values. This design uses case-based reasoning to optimize the standard temperature of coke oven. Combined with the massive measurement data and expert system optimization model, and obtains the coke oven temperatures standard value with low energy consumption and high output. The least square support vector machine model was used for the soft measurement of coke oven temperatures. Wavelet neural network was used to predict the standing fire tem- perature of coke oven. And simulation result shows that the program can find the best standard temperature setting to reduce gas and energy consumption. It provides good guidance for actual production.