An Improved Model of Ceramic Grinding Process and its Optimization by Adaptive Quantum inspired Evolutionary Algorithm

摘要

Surface grinding is a process extensively employed for removing material with specified surface finish and integrity. It is used in processing of advanced structural ceramics like silicon carbide, which are widely used in engineering applications. Ceramic grinding is a complicated process as ceramics have high hardness and tow surface toughness, which adversely affect the output quality. Optimization of the process is thus essential for maintaining desired quality while maximizing productivity with minimum cost This process has been modelled as a Continuous Nonlinear Constrained Optimization problem. Efforts have been reported to solve this problem by using Genetic Algorithms, Particle Swarms and Differential Evolution. However, most of these attempts have employed algorithms where the users J designers select parameters in the evolutionary operators. This paper improves the existing model with realistic modification and converts it into Mixed Integer Nonlinear Constrained Optimization problem. The same has been optimized by using an Adaptive Quantum inspired Evolutionary Algorithm, which is free from user selectable parameters in evolutionary operators, as they are determined adaptively. The proposed algorithm does not require mutation for maintaining diversity. Further, previous attempts have employed penalty factors or repair based techniques for handing constraints where as this effort employs feasibility rules which is again free from parameter tuning. The proposed algorithm is simple in concept, easy to use, faster and more robust than the known state of art methods available for solving such problems as shown by detailed analysis.