Study And Research Of Cold Rolling Plant Scheduling Based On Clustering Technic

Along with China's industrial development, the demand for steel products shows a growing trend. Among a wide variety of steel products, cold rolled products as a value-added products, particularly its demand. The annealing process is an important process of producing this type of steel products. Annealing process in general is divided into continuous annealing and batch annealing in two ways. Batch annealing furnace as the oldest type, technology and applications has been very mature. At the same time it has flexible operation, easy replacement of product types, production, construction, low cost, so the actual production has been more widely used.Old batch control system, a way to manually control the production, exposing complex operation, constraints, many other issues, make scheduling work into a tedious, complex and highly technical work, the method manually is difficult to ensure reasonable quantities of steel coil production scheduling. Thus leads to the transformation of Bell-type furnace system is not automated transformation, the purpose is to design a site to meet the production needs of the automation control system for batch annealing. Scheduling of work, as a process to coordinate raw materials, furnace resources, energy resources before production, affects the entire system's operating efficiency.This article describes the problems in the scheduling problem model will be followed by a detailed study of their clustering problem comes down to. The model is designed based on bisecting K-means algorithm clustering scheduling algorithm.The algorithm using C++ language in Visual C++ implementation of the simulation.Sum of the squared error(SSE) of the algorithm used to control the clustering effect and the end of the algorithm. In tests, the use of different algorithms on the obtained results are analyzed and test results indicate that improvement strategy paper, the algorithm can improve the clustering effect, thereby enhancing the productivity of actual production.