Modeling And Optimization For Separable Linear Guideway Cutting Problem With GDR Constraint

With the expanding scale of electric power industry,telecommunication industry,urban rail transit industry,automobile industry and shipbuilding industry,the machine tool industry has been developing fast,and the demand of linear guideway has increased significantly as well.Though the raw materials of linear rails have uniform specifications,customers may have various requirements for different products.It is obvious that how the splitting and cutting plans of line rail products are arranged is evident to improve the economic benefit of enterprise and to make full use of raw materials,with the process requirements satisfied.In the beginning,the thesis introduces the conception of one dimension blanking and variable size packing based on domestic and foreign literatures.By including the information of linear guideway,a detailed analysis of linear structure,characteristics and main performance index of linear cutting,process and requirements is conducted.Thus,the cutting problem of detachable linear rail is transformed to a supersize packing problem of that,and the theory and methods of solving the packing problem are summarized at the end of the discussion.The cutting method of the detachable linear rails with GDR constraints is discussed in the second part of the thesis.Under the premise of the minimum length of the rail line split,the restrictions of split position,the seams limit when splicing,the rail line length,the hole spacing,the edge distance,the cutting loss,the cutting direction,the requirements of GDR constraints and other process requirements,as well as the splitting demand of the customers,a nonlinear multi-objective integer programming model is built to achieve the goal of using least number of rail lines and producing least waste.Based on the structural characteristics and specific GDR constraints of the linear rail products,a heuristic algorithm based on the minimum waste priority and GDR constraints is designed.The algorithm begins from the characteristics of the spacing distance,the end distance of linear track and the length of the linear track.Under the premise of the GDR constraints,the algorithm takes the priority of using the shorter surplus as the greed rule,so that a small amount of raw material used as well as a high ratio of utilizing the raw material could be obtained.Genetic algorithm and artificial bee colony-genetic algorithm are developed for the multi-objective optimization model for the detachable linear rails.In terms of the splitting and cutting process of the products,a two-stage coding is designed,and the concept of scrap preference ratio is proposed to solve the two conflicting goals in the model.Then the multi-objective problem is transformed into a single objective problem.A cross variation method for splitting,cutting and the direction of cutting the products is designed,which has combined the features of the detachable linear rails,the method of chromosome coding and the heuristic algorithm.To overcome the premature phenomenon of the genetic algorithm,the artificial bee algorithm is introduced based on the existing genetic algorithm to improve the evolution process.On the basis of the initial solution of the neighborhood solution of the artificial bee algorithm,the adjacent domain solution of the suitable value is selected for crossover and variation.therefore a better offspring is gained and the convergence of the algorithm will be improved.Thus,a better cutting scheme of linear rail could be expected.Finally,the effectiveness and feasibility of the proposed model and algorithm is verified by series of experimental analysis,including the analysis of algorithm parameters experiments,the different crossover and mutation operators,the separation and analysis of experimental results of cutting experiments,the heuristic algorithm and the genetic algorithm,the particle swarm algorithm and the comparison experiment of particle swarm optimization genetic algorithm,using Excel and VBA programming.The result of comparative analysis shows that the particle swarm optimization genetic algorithm is superior to the heuristic algorithm and the genetic algorithm.