Optimization Of Rolling Schedule In Tandem Cold Mills Based On Improved SFL Algorithm

Rolling process is a complex system process, and establishing a reasonable rollingschedule is not only an important condition of guaranteeing the high quality, highefficiency and low energy consumption of the cold rolling mill operation, but also is animportant technology issue of comprehensive optimization in cold rolling production. Inrecent years, with the rapid development of industrial production, the quality andproduction of steel plate are both required increasingly high. Therefore, higher demandsfor the optimization ability of rolling schedule are put forward. With the development ofthe computer technology, a new chapter for the development of the rolling theory wasopened by the application of artificial intelligence in rolling fields. Compared with thetraditional rolling methods, the optimal control for the rolling process was achieved by theartificial intelligence algorithms which avoid endless search for the deep laws of rollingprocess in the past and simulate human brains to deal with the practical process on thebasis of the fact and data. The experience and summary in practical production can beraised to the theory level by formulating the rolling schedule. Therefore, artificialintelligence algorithm has a more extensive significance.As a new intelligent optimization algorithm, shuffle leapfrog algorithm that is basedon the heuristic search is used to solve the function optimization problems, includingmulti-function optimization problem and constrained optimization problem. Currently,shuffle leapfrog algorithm has been applied to many practical areas, and good effect hasbeen shown in many complex optimization problems. So it is a good method to put theshuffle leapfrog algorithm into the rolling schedule optimization. However, like othergroup intelligence evolutionary algorithm, shuffle leapfrog algorithm also has theshortcomings such as precocious and slow convergence speed, so a chaotic mutationleapfrog algorithm is formed through increasing variation operation on the basis of thebasic shuffle leapfrog algorithm and using Logistic chaotic sequence to construct mutationoperator.A set of cold rolling schedule optimization model in the process of high-speed rolling has been established considering the problems on the rolling load and overcomingscratches comprehensively in this paper. A good optimal precision and convergence speedhave been shown in the algorithm by regarding the relative load and overcoming scratchesas the comprehensive goal and using adaptive chaotic variation leapfrog algorithm tooptimize the model. To achieve better forecast results, the neural network model of rollingforce which is the most important rolling parameter has been established, and adaptivegenetic algorithm is also applied to optimize the model. The simulation practice showsthat the satisfactory results can be obtained by the optimal rolling schedule.