【摘要】 为解决热轧过程中电能消耗和轧辊磨损的多目标热轧计划优化问题，建立以最小化耗电成本和板坯间规格跳跃惩罚值为目标的优化模型，为了高效求解该优化模型，提出了一种改进遗传算法(IGA)。该算法采用嵌入启发式规则的初始化策略以提高初始种群质量；基于sigmoid函数的自适应交叉变异操作以提高算法解集的广泛性和多样性；将精英选择与轮盘赌策略结合的选择策略以提高种群多样性并保留进化过程中的优质个体。为了测试改进算法的性能，以某钢厂实际生产数据为例进行求解，设计该算法与传统遗传算法、原始人工蜂群算法、模拟退火算法进行对比，从目标值均值、优化率和变异系数对实验结果进行分析，结果表明改进策略有效提升了遗传算法的探索性和稳定性；与传统遗传算法相比，所提出的改进算法优化率提高了3.61%,耗电成本降低了2.87%,即改进后的算法能有效节约能源成本，实现热轧计划优化问题的高效求解。更多还原

【Abstract】 In order to solve the multi-objective hot rolling planning optimization problem of power consumption and roll wear in the hot rolling process, an optimization model is established to minimize power consumption cost and the penalty value of specification jump between slabs. In order to solve the optimization model efficiently, an improved genetic algorithm(IGA) is proposed. The algorithm adopts an initialization strategy embedded with heuristic rules to improve the quality of the initial population. The adaptive crossover and mutation operation based on the sigmoid function to enhance the universality and diversity of the algorithm solution set. A selection strategy that combines elite selection with a roulette strategy to improve population diversity and retain quality individuals in the evolutionary process. To evaluate the enhanced algorithm′s performance, real production data from a steel mill was used as an example to solve the algorithm. The original artificial bee colony method, the simulated annealing algorithm, and the conventional genetic algorithm were all used in the construction of the algorithm. Utilizing the mean goal value, optimization rate, and coefficient of variance, the experimental findings were examined. The suggested improved algorithm outperforms the conventional genetic algorithm in terms of optimization rate by 3.61% and power consumption cost by 2.87%. In other words, the improved algorithm effectively reduces energy costs and achieves an efficient solution to the hot rolling plan optimization problem.更多还原