Two-layer Optimization Of Process Planning By Genetic Algorithms

Optimization of process planning is the important method to ensure product quality, to improve efficiency and to reduce cost. Optimization algorithm for process planning is related with many elements such as machining method, operation sequencing, cutting parameter, machining tool and cutting tool etc, which are affected one another. A uniform optimization model of process planning is brought forward to gain the global optimum in entire solution space for process planning using improved genetic algorithm, and the process planning is represented by the gene group-based special chromosome-encoding method with double-layer structure. Operation sequencing, machine tool-selecting, cutting tool-selecting and cutting parameters in each operation are optimized as the whole. At the same time, the gene-group-based encoding schema and two-layer optimization strategy are presented. The optimal process planning is educed by two-layer optimization strategy-based genetic computation.The dissertation introduces the project background, source and aim, and summarizes status quo of CAPP and method for optimization of process planning. The main tasks are indicated.In order to compute punishing value resulting from operation sequencing violating constraint, the constraint relation of machining operation sequence is analyzed, and the model is built for it. The uniformed process resource models, including organizing relation model and object model, are established in uniform mark language, and the relation pattern is design, which is convenient for the second layer mutation operator of the evolutional computation. As the result, the recomposing-oriented process resource database is constituted. The cooperation management-oriented process resource mark language is developed to boost up fitness of the process resource database by extensible mark language.The matrix for machining tool-cutting tool-mathematics model is advanced, convenient for optimization of process planning. The mathematics model for machining operation is formulated. The evaluating system for process planning, considering such criterion as cost, time, profit and quality, is created by analytical hierarchy process, which is based to compute the fitness of process planning chromosome.After basic genetic algorithm is introduced, an improved genetic algorithm is projected, in according to the multi-layer no-linear combination optimization trait of process optimization planning and the algorithm is particularly dissertated. The improved genetic algorithm based method for two-layer process optimization planning is advanced.The prototype system that realizes the research contents of the dissertation are given, and the prototype system is applied to a practical process planning case. Therefore, the feasibility and the effectiveness of research results are validated. In the end of the dissertation, conclusions have been drawn and the future directions in the field are given.