Order-grouping Optimization Research In MES Of Steel Pipes Enterprise

With the rapid development of national economy in recently years, the one-dimensional cutting stock problem occurs in many industry areas. For example, cloth cutting in clothing manufacturing industry, leather segmentation in leather products manufacturing industry, etc. The cutting of meta material plays an import role in cutting stock problems because a medium-sized manufacturing enterprise whose main products is metal material will need to cut thousands tons of various metal materials. Even if we can only increase the utilization of raw materials by 1%, the material-saving effect is also very impressive and it will bring huge benefits to enterprises.The steel-pipe cutting stock problem that the thesis concentrates on is a special one-dimensional case, which can exactly be called a two-stage cutting stock problem. And the whole process is divided into two stages:the hot rolling and cold rolling. Let me make a brief introduction about them:the hot rolling is defined as that steel pipes will be cut into the middle length pipe (with length limit) after the stretch reducing on the production floor area; the cold rolling is that the middle length steel pipe will be cut for the supply need of orders.Firstly, in this paper, we study the one-dimensional cutting stock problems in great detail and put forward an algorithm that can generate cutting patterns automatically by the computer. Then we establish a mathematical model based on these patterns. Since the problem is a typical linear programming problem, we use Gaussian Elimination Methods to solve it.Secondly, based on comparison between the steel-pipe CSP and classic CSP, we set up the mathematical programming model with objective function of minimizing input. Lingo has unparalleled advantages in solving linear programming and nonlinear programming, moreover, its syntax is simple and easy to grasp. So we implement an optimization software to solve the fixed order-grouping(ie, steel-pipe cutting) problems with a hybrid programming approach which combines Java and Lingo. The optimization software is user-friendly and proved effective by a large number of experimental data.Finally, we propose fixed order-grouping optimization algorithm. Based on industrial background, we first establish a two-stage model:in the first stage, we derive the intermediate length of steel-pipe with the backstepping approach in mathematical logic according to the length and corresponding number of orders; the second stage we use a heuristic algorithm to solve, which is proved effective and fast through. By randomly selecting groups of large-scale data to test, we can quickly obtain the optimal solution or approximate optimal solution.