论文部分内容阅读
针对多目标优化算法中综合目标函数权值难以确定的问题,通过对偏好的数学量化,采用数值分析的方法,构造了6种不同的偏好函数,建立了物理规划(physical programming)数学模型,然后以遗传算法为寻优工具,实现一种更加灵活更加适合于工程技术人员的交互式多目标优化算法。结合某冷轧厂实际的轧制规程优化过程,选取等功率裕量、轧制总能耗及各机架打滑因子为目标函数,运用基于遗传算法的PP进行优化计算。结果表明,优化后的轧制规程很好地实现了各机架等功率负荷分配,降低了打滑出现的概率,大大提高了板材表面质量和成品成材率。
Aiming at the difficulty of determining the weight of the comprehensive objective function in the multi-objective optimization algorithm, six different preference functions are constructed by numerical analysis of the preferred mathematical quantification and a mathematical model of physical programming is established. Then, Taking the genetic algorithm as the optimization tool, an interactive multi-objective optimization algorithm that is more flexible and more suitable for engineers and technicians is realized. Combined with the actual rolling schedule optimization process in a cold rolling mill, the objective function of power margin, total rolling energy consumption and each frame slippage factor was selected. The PP based on genetic algorithm was optimized. The results show that the optimized rolling schedule can well realize the power load distribution of all racks, reduce the probability of slippage and greatly improve the surface quality of sheet metal and finished product yield rate.