Load Distribution Optimization Of Hot Rolling Based On Ant Colony Algorithm

Flat steel hot rolling is one of the most important processes of steel production, and set-up calculation of Hot Strip Mills is the basic work of continuous rolling operation and the most important problem of it is press load distribution of each stand. At present, the traditional method that relies on experience to distribute rolling load cannot meet the demand of modern rolling production. And make a value rolling load distribution is the fundamental guarantee of giving full play to the rolling equipment capacity, energy conservation, and fulfilling quality requirement of strip product. Therefore, it has important project application value to use artificial intelligence technology to study hot rolling load distribution.Energy conservation is an important indicator of steel rolling control, and it is also the prefatory topic of study on strip rolling technology. The reasonable load distribution can make the control of equipment energy consumption easier to implement. In this paper, take hot strip mill in a domestic factory as object of study, the optimized research of hot rolling load distribution and set-up calculation is aim to reduce energy consumption, and on the basic of traditional mathematical model, and use ant colony optimization. The main content includes:(1) Development condition of hot continuous rolling band steel, and main mathematical model of hot continuous rolling set-up calculation has been introduced in detail.(2) Analysis and research of AS and MMAS has been made, and rely on oliver30, the influence of parameter in ant colony algorithm to algorithm has been verified and analyzed, and the optimal algorithm parameter has been determined.(3) Base on the optimization feature of ant colony optimization, the method of applying ant colony optimization to the optimization of rolling load distribution has been designed, and use C++language to write simulation example. Based on experience load distribution, and use AS and MMAS to optimize load distribution, then contrast the optimize results of AS and MMAS. Simulation results show that convergence rate of MMAS is slower, but the global convergence performance is better than AS. (4) To aim at hot strip mill in domestic factory, and base on original RAS platform, use VC++2008to design optimizer of load distribution, then take field testing. At last, compare the result of load distribution that has been optimized by MMAS and original energy consumption curve method. Field measurement showed that the result of load distribution that has been optimized by MMAS, the outgoing thickness, reduction ratio and distribution of rolling force of each stand is more reasonable. It can ensure that the first few stands can make full use of equipment capacity and also can satisfy that the requirement of decrease rolling force progressively due to the plate shape, thick precision and performance index of last few stands. The result of load distribution is better than that made by energy consumption curve method.