Application On The Path Optimization Problem With Intelligence Fusion Algorithm

According to the quality information management system in CIMS (computer integratedmanufacture system) machining processing under CIMS, The architecture of qualitymanagement system of machining processing, which decomposes system functional actionlayer-by-layer with IDEF0method is designed. In this design processing, in order to keep dataconsistency and integrity and eliminate data redundant, database concept design is conductedby using IDEF1x method to establish the system information model, and the conceptconstruction model is developed on this basis. Aiming at path optimization problem inintegrated system, basis path information element according to the math description isproposed, by using TSP math model in graph theory, the math model of path optimization isbuilt, and construct the intelligent fusion optimization algorithm, which is applied in thegroup hole processing.At present, intelligent fusion optimization algorithm, which shows nice effect on pathoptimization problem, and has made rapid spread and application in many engineering fields.However, with the improve of complexity and scale of path optimization problem, it is moreand more difficult to get a satisfying result by using some single optimization algorithm ordepending on the improvement of some parameters and indexes. It has become a better choiceto solve the problem together, by complementarily fusing this algorithm with mature method.In this paper, fusing local search algorithm, simulated annealing algorithm with antcolony algorithm properly, intelligent fusion optimization algorithm is built. In the stage ofoptimized initial solution, by using3-opt algorithm and modified LK algorithm, it has speedup the convergence and avoid the low efficiency and local optimum. By using local search,optimized results, which will be treated as initial value of pheromone, are fast generated, thusimproving searching speed of prophase ants. The shortage of occurred premature convergencelocal optimal value and low convergence speed in revolution processing of traditionalalgorithm, and Metropolis probability acceptance criterion is added. Taking group holeprocessing for example, cutting path is optimized in machining processing, which realizes thepoint position optimization in CAD/CAM multi-feature processing, shortest processing pathand shortened idle waking-cutter time, which also greatly improves the processing precisionand efficiency.