基于热力学混合遗传算法的加热炉炉温优化研究

针对加热炉生产过程中钢坯入炉温度、规格尺寸、钢坯种类等生产工况经常会发生改变,导致基本遗传算法存在早熟等现象,提出一种基于热力学的混合遗传算法.基于钢坯加热过程的机理模型,建立了钢坯温度预报模型,依据加热炉工艺生产要求,建立了加热炉炉温优化模型.为了提高遗传算法的求解精度和计算效率,在遗传算法交叉算子设计过程中加入内能、熵和自由能的思想,改进了传统遗传算法;同时在经典的遗传算法基础上加入模拟退火算法构成了基于热力学的混合遗传算法,并用于求解加热炉炉温优化问题,克服了传统遗传算法的不足.实验结果表明,该方法能够有效地求解加热炉炉温优化问题,是可行的、有效的.

Heat furnace temperature optimal problem based on hybrid genetic algorithm

In view of the initial temperature, billet size, species and other production conditions changing, the basic genetic algorithm has premature phenomenon.A hybrid genetic algorithm which is based on the thermodynamic is proposed. Based on billet heating mechanism, a slab temperature prediction model is established. Then, according to the furnace production process, a temperature optimal model is established. In order to improve solution accuracy and computational efficiency, a crossover process in genetic algorithm is enclosed by internal energy, entropy and free energy of thought,to improve the traditional genetic algorithm.While a classical simulated annealing algorithm is combined with genetic algorithm,a hybrid genetic algorithm is used to solve the heating furnace temperature thermodynamic optimization problems,which can overcome the shortcomings of the traditional method. Experimental result shows that this method can effectively solve the heating furnace temperature optimization problems, so this method is feasible and effective.