Study On Models And Algorithms Of Production Planning And Slab-yard Optimization Management In Hot-rolling

Steelmaking-continuous casting-hot rolling in iron and steel plant includes continuous and spiccato production process, and it is a typical hybrid system. Because the three working procedures of steelmaking, continuous casting and hot rolling run sequentially, integrated production scheduling needs to consider not only materials circulation and resource balance, but also time and energy balance under the production circumstance of high temperature. Since 1980s, steelmaking-continuous casting-hot rolling integrated production is studied and applied by overseas large iron and steel plant to save energy resources and reduce wastage, and steelmaking-continuous casting-hot rolling integrated production planning and scheduling become problems urgent to be solved. This dissertation has studied production planning of hot rolling and slab-yard under steelmaking-continuous casting-hot rolling integratied production planning system, and has suggested different intelligent algorithms to solve the models proposed. This dissertation has mainly carried on the following research.The model system of steelmaking-continuous casting-hot rolling integrated production planning is studied, and an integrated production planning model method is suggested. It decomposes the optimization problem of integrated production planning into five local optimization problems. Each local optimization system includes several models, can produce its production plan by using multi-model cooperation. The integrated production plan is produced by the five local optimization systems with the mode of multi-systems cooperation.The day production planning model and algorithm of hot rolling based on Just-In-Time are built. Now, hot rolling plant faces various market demands such as large variety, small quantity, high quality, low price, rigorous consignment date etc., this dissertation suggests a production planning optimization model of hot rolling based on the Just-In-Time idea and uses day as time unit. An improved mixed genetic algorithm is designed to solve the model, which introduces a new-layered addressing coding method based on natural number, uses a selective operator constituted of roulette and tournament selection, dynamically adjusts the probability coefficients and amended every result with intelligent heuristic methods.According to rolling planning problem, a rolling plan VRP model and algorithm are suggested. The model considers not only the slabs' bounce on gauge and hardness, obverse and inverse bounce on width, and the rolling length constraint of the slabs with same width, but also the reasonable arranging of warm-up materials and staple materials, and it can prudece whole roll plans which are "tortoise shell" shape and meet with the hot rolling rules. In order to solve the model, an immune algorithm based on partheno-genetic operators and expert system methods is suggested.Considering operation flow of slab entering slab-yard, the slab location decision optimal model and algorithm are suggested. The model considers several rules of slab location and pile position selection and is a general optimal operation to a slab lot of entering slab-yard. An adaptive chaos genetic algorithm is used to solve the model. The algorithm uses natural number coding method with dynamically adjustment for the probability coefficients of crossover and mutation, and uses chaos optimization method as the mutation operator.Slab discharge planning models and algorithm are proposed. The slab discharge planning is decomposed as two combinatorial optimization problems: slab discharge optimization decision and optimal turned-out slab pile. A slab discharge optimization decision model is proposed for the first problem and another optimal turned-out slab pile decision model is proposed for the second problem. The slab discharge planning can be accomplished by the two models cooperation. A discrete particle swarm optimization (DPSO) algorithm is designed to solve the models suggested. The DPSO algorithm uses a particle's value selecting mode and velocity changing pattern for the models.Using the object-oriented programming design method, hot rolling production scheduling and slab-yard optimization management system is designed and carried out, which is the subsystem of steelmaking-continuous casting-hot rolling integrated production scheduling simulation system, and bases on the models and algorithms suggested in this dissertation. The simulation application result shows the optimal scheduling method improves production efficiency and automation management.