DEVELOPMENT OF AN ENHANCED FEATURE RECOGNITION SYSTEM AND ITS APPLICATION FOR OPTIMIZIED PROCESS PLANNING OF SHEET METAL BENDING

Abstract

The efficient process planning of the V-bending processes involves the determination of a feasible sequence and tool stages of the bending tasks to achieve the final desired product shape. The feasibility of such a sequence is materialized by the absence of collision during V-bending processes. According to the interference nature of the tasks of the V-bending process planning, it is considered as a constrained combinatorial optimization problem. In this thesis, the proposed Computer Aided Process Planning (CAPP) system uses the genetic algorithm as an optimization search algorithm to produce near optimal process plans. The proposed CAPP system includes three modules which are feature recognition module, collision detection module, and genetic algorithm optimization module. In the proposed system, the optimization algorithm is linked with the recognized features of the bent workpieces and the relations between the bend lines which could guide the search to converge to the near optimal process plan in minimum number of generations.