基于仿生算法的全废渣水泥生料配料优化研究

Research on optimization of raw material batching of all-waste slag cement based on bionic algorithm

针对在生料配比过程中出现原料成分复杂、波动大而人工计算配料法无法满足生产需求的问题,基于固废原料的特征开发出以非支配排序算法为代表的全废渣水泥生料配料优化方法,输入历史进料数据后产生初始种群,经过巧妙设置的目标函数进行遗传迭代计算,直至满足终止条件输出最优解.结果表明:基于仿生算法的全废渣水泥生料配料优化方法可以满足均衡地接近目标三率值,并优先消纳库存压力较大的固废;相比传统的遗传算法,其对电石渣的消纳量提升了23.1%,计算时间缩短至30.8 s,计算速度提升了74%,迭代次数降低了73%.

To address the issues that the raw material composition is complex and fluctuates greatly in the raw material batching process and the manual calculation method cannot meet the production demand,an all-waste slag cement raw material batching optimization method was developed as represented by a non-dominated ranking algorithm based on the characteristics of solid waste raw materials.The initial population was generated after using the historical feeding data as input,and the genetic iterative calculation was carried out through the cleverly set objective function until the termination criteria is satisfied and an optimal solution was obtained.Results show that the all-waste slag cement raw material batching optimization method based on the bionic algorithm can approach the target triple modulus in a balanced way and give priority to the consumption of solid waste with high inventory pressure;compared with the traditional genetic algorithm,the consumption of calcium carbide slag is improved by 23.1%,the computation time is shortened to 30.8 s,the calculation speed is improved by 74%,and the number of iterations is reduced by 73%.