对于焦炉加热这一复杂工业过程,提出一种包括协调层和优化控制层的多工况火道温度优化控制方法。协调层根据对焦炉加热过程工艺参数的分析,采用多信息融合的二次决策方法,由荒煤气的温度识别焦炉加热过程的实时工况,针对不同的工况选择合适的优化控制模型。基于工况分析,在优化控制层采用一种基于自适应遗传算法的多目标模糊优化控制方法,针对不同工况下的模糊控制器量化因子和比例因子调节困难的问题,采用精英保留和赌盘算法相结合的选择策略,以及具有自适应交叉概率和变异概率的遗传算法对模糊优化控制模型的参数寻优,有效地提高了遗传算法的全局搜索能力和收敛速度,并且通过对控制精度、能量消耗和调节时间等各项指标适当加权,构造适应度函数,使优化后的模糊控制模型达到满意程度。采用具有多工况火道温度智能优化控制结构的方法取得了良好的控制效果,为焦炉加热过程的优化控制问题的解决提供了一条新的途径。
An optimization control method was presented for the flue temperature in multi-operative modes in the coordinating layer and the optimizing control layer for coke oven heating process. At the coordinating layer, the real-time operative mode of the heating process which was identified by the raw gas temperature, was acquired through the information fusion method of second decision. Proper optimization control model was chosen according to different operative model. At the optimizing layer, a multi-objective optimization method for optimizing parameters based on adaptive genetic algorithms was proposed to deal with the difficulties in tuning parameters of the fuzzy control model. By using elitist strategy and roulette wheel algorithm in selection adaptive crossover and mutation probabilities, the global searching ability and the convergence speed of the genetic algorithms were significantly improved. Through properly weighting the terms including control errors, energy consumption and tuning time to construct the fitness function, the fuzzy control model can be optimized to satisfaction. Obvious effect of this method was obtained in the application and a new approach for the optimization control of coke oven was provided.