Research On Round Billet Design Problem Of Hot-rolled Seamless Steel Pipes

In this dissertation,the definition of round billet design problem(RBDP)of hot-rolled seamless steel pipes is proposed.A systematic study is carried out for the RBDP with single round billet weight,multiple round billet weights,process constriants and constraints of orders,round billets and process.The main contents and innovations are as follows.(1)From the hot rolling process of seamless steel pipes,a definition of RBDP was proposed firstly based on the background,features and constraints of the problem.For the most fundamental RBDP,a mixed integer model was established to minimize the number of round billets used.The most fundamental RBDP is proved to be NP-hard by reducing the 3-partition problem to it,and the tight lower bound of the RBDP could be calculated.According to the features of the RBDP,two rules named order assignment rule(OAR)and order reassignment rule(ORR)were presented based on the minimum allocation weight of orders,and then a two-stage heuristic algorithm based on greedy strategy(HSGA)was proposed to solve the problem.Experiments proved HSGA was both effective and efficient.(2)For the RBDP with the constraints of multiple round billet weights,a multi-objective mixed integer model was fonnulated to minimize the leftover of round billets and the number of round billets used at the same time.Based on the minimum allocation weight of orders,a method to obtain all the sub-order combination schemes for orders was given.To solve the problem,a multi-objective algorithm named INSGAII was proposed with novel encoding and decoding operations,recombination and mutation operations and the improved crowded intensive sorting strategy.Simulation results proved INSGAII could achieve good Pareto solutions with good stability.(3)Considering the process constraints,the RBDP could be seen as a two level nesting problem:the outer level is the sub-problem to decide the number and weight of short round billets to produce order pipes;the inner level is the sub-problem to allocate all the short round billets generated in the outer level to the round billets.Based on that,a mathematical model is established to minimize the number of round billets used and the number of short round billets genertated.As the solution space of the outer layer sub-problem can be calculated by the known conditions,a strategy based on prediction and dynamic programming is proposed,and based on that,a heuristic algorithm is presented,which can solve large-scale RBDP in a short period of time with good stability.(4)For the RBDP with the constraints of order weight and pipe length in interval values and multi round billet weights,considering the process constraints as well,a model is established to minimize the leftover weight of round billets used and the weight of unplanned pipes based on the analysis of constraints of orders,billets and process.By analyzing the relations among variables,the domain of RBDP is reduced and the RBDP model is then simplified.To solve the problem,a heuristic algorithm based on dynamic short round billet selection rules,order relaxation strategy and greedy strategy.Experimental results show that the algorithm could meet the special requirements well in practical production.