焦炉集气管压力设定值多目标优化研究

Multi-objective Optimization Simulation of Setpoint Pressure of Coke Oven Collectors

针对某焦化厂焦炉的集气管压力设定值主要是由工艺工程师人为给出的生产现状,提出集气管压力设定值多目标优化的研究方案。利用焦化厂海量的炼焦历史生产数据,综合考虑了炼焦能耗,焦炭产量及质量三个目标,采取了基于K-均值聚类的径向基神经网络来建立其与集气管压力的关联模型,并通过基于差分进化的粒子群的多目标优化方法来获取集气管压力设定值,通过仿真研究验证了方案的有效性。结果表明,上述优化方案可以给出最佳集气管压力设定值在保证质量的前提下,使能耗降低,产量提高,满足现场需求,可以为实际生产提供操作指导,为压力控制系统奠定了基础。

Aiming at the problem that the setting value of the collector pressure for the coke oven of a coking plant is mainly based on the production status given by the process engineer, a research for the multi-objective optimization of the set pressure of the collector is proposed in the paper. Using the coking plant’s massive historical coking production data, taking into account the three goals of coking energy consumption, coke production and quality, a radial basis neural network based on K-means clustering was used to establish a correlation model with collector pressure. The multi-objective optimization method for particle swarm optimization based on differential evolution was used to obtain the set pressure of the gas collector. The effectiveness of the proposed scheme was verified through simulation. The results show that the optimization scheme can give the best collector pressure setting value under the premise of guaranteeing quality, reduce the energy consumption, increase the output, meet the on-site requirements, and provide operational guidance for actual production, which lay a foundation for the pressure control system. basis.