Optimization of secondary cooling water distribution system for slab continuous casting based on self-heuristic algorithm
This study uses the Q355B steel production process of a steel plant's No.10 slab continu-ous caster as the research background and systematically compares the application effects of various heuristic algorithms such as particle swarm optimization(PSO),genetic algorithm(GA),simulated annealing(SA),teaching-learning-based optimization(TLBO),and particle swarm optimization-genetic algorithm(PSO-GA)in optimizing the secondary cooling water system of continuous casting.The research formulated multiple optimization targets based on metallurgical standards to optimize the continuous casting process and used a heat transfer numerical model to simulate the solidification and cooling of the slab.A comparison of different algorithms shows that while the PSO algorithm has high solving efficiency and fast convergence,it exhibits considerable fluctuation and low stability.The GA algorithm demonstrates higher stability but slower convergence.The SA algorithm,with simple param-eter adjustments,offers the fastest computing speed and good stability but slow convergence and low precision.The TLBO algorithm,despite its complex structure which results in the longest computation time,achieves high stability,rapid convergence,and high precision.The PSO-GA hybrid algorithm maintains rapid convergence while significantly enhancing global search capabilities,substantially im-proving optimization stability and accuracy.In terms of optimization effects,all algorithms successfully improved the uniformity of slab cooling and temperature distribution.These results not only confirm the potential application of heuristic algorithms in continuous casting technology optimization but also provide a theoretical reference for further digital research and practical application of algorithms in secondary cooling processes of continuous casting.
slab continuous castingsecondary coolingheuristic algorithmwater distribution system optimizationnumerical modeling
NETL
NSTL
万方数据
