Research On The Modeling And Optimization Algorithms For The Slab Design And The Hot Rolling Production Planning Of Steel Plate

Production planning and scheduling are identified as key contents of the intelligent management of steel enterprises and play important roles in the process when the iron and steel industry achieves intelligent manufacturing.Steel plate is a type of very important material widely used in construction,engineering machinery,national defense,and other industries.Commonly,customer orders for steel plate are characterized by multiple sizes and small batches,which makes the production planning of steel plate more difficult.The main difficulties lie in how to design slabs and hot rolling units so as to match the large-scale and batch production mode of steel enterprises.This dissertation focuses on the research on the modeling and optimization algorithms for the slab design problem,the cold charge hot rolling production planning problem,and the hot rolling production planning problem considering hot charge processes,all of which take the characteristics of steel plate orders and production into account.The main research findings and conclusions are given as follows:For the slab design problem of steel plate(SDPSP),its mixed integer programming model is constructed,and the hybrid optimization algorithm based on tabu search and column generation technology(TSCGHOA)is proposed.Slab design is to design slabs according to the product requirements of the customer's orders.The SDPSP involves combining child plates and determining the sizes of mother plates and slabs,and it is formulated as a variant of the cutting stock problem that considers making decisions on the sizes of the mother plate and the slab.The size flexibilities of the mother plate and slab are transformed into constraints,and the optimization model with the objective minimizing the surplus material and material loss in the manufacturing process is proposed.The TSCGHOA is employed to solve it.Multiple experimental instances are generated to test the performance of TSCGHOA,and the results show that the proposed TSCGHOA is able to generate good solutions for the slab design problem where multiple sizes of slab sections are involved.For the cold charge hot rolling production planning problem of steel plate(HRPCC),two solution approaches are proposed based on constraint programming and mathematical programming,respectively,and their performance is compared.The hot rolling production planning problem is to group and sort slabs(orders)from the slab set(order set)to generate one or several rolling units.The hot rolling production planning problem of steel plate requires to design a suitable method for evaluating the influence of the combination and sorting of slabs on roll wear.Based on the designed method,the HRPCC which generates multiple rolling units is described as a variant of the team orienteering problem with production constraints specific to steel plate.The constraint satisfaction optimization model and the mixed integer programming model for the HRPCC are proposed,respectively,and the objective is to maximize the total standard rolling length of scheduled slabs.The two models are solved by the software CP Optimizer and Gurobi,respectively.Multiple experimental instances are generated to test their performance.The results show that the quality of the solutions obtained by CP Optimizer is increased on average by 31.14% compared with Gurobi.For the hot rolling production planning problem of the steel plate production considering hot charge processes(HRPHC),the bi-objective mixed integer programming model taking the time-dependent heat loss into account is built,and the multi-objective evolutionary algorithm based on the decomposition technique and iterative local search(HMOEA/D)is developed.Compared with the traditional cold charge process,hot charge processes have great advantages in energy saving and consumption reduction.It is necessary for HRPHC to consider both the production efficiency of the hot rolling mill and the energy saving of the heating furnace.Based on the method proposed in the previous chapter,the HRPHC to generate one rolling unit is formulated as an orienteering problem.Taking into account cold slabs and hot slabs which include the slabs that may not yet be produced,a bi-objective optimization model is proposed and its objectives are to maximize the collected potential heat contained in hot slabs and the total standard rolling length of scheduled slabs.The HMOEA/D is developed to deal with it.Experimental instances are constructed to test the performance of the algorithm.The HMOEA/D is compared with the non-dominated sorting genetic algorithm and the multi-objective ant colony algorithm,and the results show that the HMOEA/D is able to obtain Pareto fronts with good convergence and distribution on more than 90% of the instances.Based on the above proposed models and solving algorithms,the slab design and hot rolling production planning assistant decision system for steel plate is developed.According to the system requirements,it is mainly divided into three modules for SDPSP,HRPCC,and HRPHC,respectively.The system has a human-computer interaction interface,which enables users to finish planning by clicking buttons with the mouse.It is able to screen and manage the input data,set the planning process parameters,and display and adjust the calculation results.In order to solve the difficult problems in the production planning of steel plate,the optimization models for the SDPSP,HRPCC,and HRPHC are proposed,and corresponding solution methods are developed in this dissertation.Based on these models and algorithms,the slab design and hot rolling production planning assistant decision system for steel plate is developed.The approaches involved in this research provide new solutions for the modeling and solving of the multi-constraint and multi-objective production planning problems arising from the steel industry.In addition,a decision-making tool is presented for steel plate production arrangements.