Research On The Optimization Of Load Distribution Of Mill Based On Hybrid Genetic Algorithm

The most essential work when rolling band steel is the setting calculation for schedule, while the core problem for setting calculation is to allocate the reduction between each pass per body framework and determine the actual rolling thickness for each framework, furthermore allocating reduction load for the according framework in specific pass. Reasonable load allocation makes roller control much easier and achieves better control. As a result, load allocation when rolling has much great significance.This article carries out optimization research on load allocation for hot continuous rolling using combined genetic algorithm, taking load allocation for continuous rolling unit set with seven frameworks as example, based on conventional mathematics models. The main content for this article are as follows:It introduces development overview of load allocation when rolling band steel and lists the popular mathematics model related when calculating settings for hot continuous rolling.It also introduces application of optimization in metallurgy and analyzes GA and sequential quadratic programming, pointing out their respective advantages and disadvantages. Based on it, a brand new hybrid GA has been designed, which has both advantages of these two methods. This algorithm designs new genetic operator and proposes feasible proposal, aiming at accelerating local search and improving global search. Simulation experiment is carried out for the new algorithm using typical testing function and compares the results to the GA. The result demonstrated that convergence speed and global search performance precedes the GA.From the simulation experiment, it can be learnt that load allocation produced using combined GA optimization including outlet thickness, reduction rate, rolling force allocation is more reasonable. It can not only ensure that first few frameworks can sustain the possible largest reduction, but also satisfy requirement that last few frameworks sustain decremented rolling force due to shape and thickness precision and performance norms. The improved result is much better load allocation based on experience.