Modeling And Solving Of Steel-making Continuous-casting Scheduling Optimization Problem

Iron and steel industry is an important industry in our country, which is important topromote China’s economic development. Steel making and continuous casting is animportant stage in the production process, and efficient scheduling scheme can effectivelyreduce production costs, reduce energy consumption and improve steel quality andproduction efficiency. However, steel making and continuous casting productionscheduling is a complex mix of no-wait flow shop scheduling combinatorial optimizationproblems, and also NP-hard problem。There doesn’t exist an optimal solution inpolynomial time. With complex constraints and many uncertain dynamic events, and highreal-time requirements, the scheduling plan is very difficult to make. Therefore, theproblem of steel making and continuous casting production scheduling research hasimportant theoretical significance and practical value.For small-scale steel mill’s single machine scheduling problems， Since thetemperature is particularly high during the process of steel-making continuous-casting, ascheduling model in which the buffers are quantity-limited and time-limited is created.Besides considering the situation of continuous casting, the time for changing the tundishand the limited waiting time before continuous-casting are also taken into consideration ashard constraints to guarantee the quality of steel embryo. A discrete Artificial Bee Colonyalgorithm combined with several neighborhood search methods is proposed to solve thescheduling problem of steel-making continuous-casting. In this algorithm, a modificationcount is used to show one solution’s status as well as switch search methods, whichaccelerates the process of searching excellent neighborhoods. Experimental tests on ainstance derived from a real steel-making factory and several randomly generatedinstances validate the effectiveness of the proposed algorithm.For large-scale steel mill’s multiple machine scheduling problems，According to thereal-world process environment, a steelmaking and continuous casting scheduling modelwith multi-process is built. Conversion rules are added to simplify the problem. Theintroduction of variable neighborhood search is used to adjust the machine sequences anddiscrete Artificial Bee Colony algorithm is used to adjust the job sequences，which makesparallel problems serialized。Finally by analyzing the result of tests with several instances，a conclusion is made between manual scheduling, discrete Artificial Bee Colony algorithm, discrete Artificial Bee Colony algorithm with VNS and Genetic Algorithm to verify thefeasibility and effectiveness of the proposed algorithm.At last, based on the previous chapters, a conclusion of the full paper is made, whichmay be a reference for subsequent researchers.