【作者】 王景胜；

【导师】 王洪瑞；

【作者基本信息】 燕山大学 ， 模式识别与智能系统， 2012， 硕士

【摘要】 轧制规程是冷连轧工艺研究的主要内容，合理的轧制规程是板带轧制精度与板型良好的根本保障，更是冷轧设备长期可靠稳定运行的重要保证。冷轧过程复杂，采用何种优化方法和轧制策略对轧制规程进行优化，都会对最终轧制规程的合理性产生影响。同时，传统方法的效果不佳，人工智能算法的不成熟，制约着轧制规程优化结果的最优性。本文从优化方法入手，分别从单目标与多目标的角度对轧制规程进行设计。首先，从优化方法出发，分析各种优化方法特点，确定了将具有快速收敛性的粒子群算法作为轧制规程的优化方法，通过与混沌搜索结合，给出了协同优化算法，通过对测试函数进行仿真实验，协同优化算法利用混沌序列的均匀遍历性很好的克服了粒子群算法早熟收敛的缺点。然后，根据1250mm八辊五机架冷连轧机的技术参数与设计特点，以等相对轧制功率分配作为目标对轧制规程进行优化设计，并采用协同优化算法作为优化方法，对轧制功率进行均衡分配。最后，从实际生产中多方面要求考虑，通过对帕累托最优理论的研究，阐述了处理多目标问题的优化方法，进而给出了基于协同优化算法的多目标轧制规程优化设计。通过帕累托最优得到的轧制规程，更加符合实际生产的需求。更多还原

【Abstract】 The rolling schedule is the main content of the cold rolling process, and reasonablerolling schedule is not only the fundamental guarantee of strip rolling accuracy and goodplate shape, but also the significant support of long-term reliable and stable operation ofcold rolling mill. In view of the complexity of the cold rolling process, the optimizationmethods and the rolling strategies chosen to optimize the rolling schedule will influencethe rationality of the final rolling schedule. At the same time, traditional methods ofrolling schedule optimization are ineffective, while the artificial intelligence algorithmsare immature. Both of them lead to a limited optimization of rolling schedule optimizationresults. In review of the optimization method, this paper designs the rolling schedule fromthe aspects of single objective and multi-objective respectively.Firstly, starting from the optimization method, a variety of optimization methods wasanalyzed, and the particle swarm optimization with the rapid convergence is chosen as theoptimization method of rolling schedule. Then, the collaborative optimization algorithm isput forward through the combination of particle swarm optimization and chaotic search.Simulation results of the test functions prove that collaborative optimization algorithm canovercome the premature convergence of particle swarm optimization by the uniformergodicity of the chaotic sequence.Then, considering the technical parameters and design features of the 1250mmeight-roller five-stand tandem cold rolling mill, a rolling schedule optimization is designedwith equal relative rolling power distribution as the target, using collaborativeoptimization algorithm as the optimization method. Experiment results show that therolling power achieves a balanced distribution after optimization.Finally, taking the many requirements from the actual production and study of paretooptimality theory into account, the dissertation explains the optimization methods dealingwith multi-objective problem, and then puts forward a multi-objective rolling scheduleoptimization design based on collaborative optimization algorithm. The rolling scheduleobtained through pareto optimality is more in line with actual production requirements.更多还原