A Web-Based Platform for Intelligent Instrument Design Using Improved Genetic Algorithm

摘要

The design of new products is a creative work based on designer's knowledge or experience. This paper develops a web-based design platform for intelligent instrument with the technology of Java and web database. It aims at assisting designers in product decision-making with the selection of modules and offering near-optimal solutions of product design scheme that meets user requirement. An improved genetic algorithm combined with simulated annealing algorithm is proposed to accomplish the selection of module more effectively. In order to avoid prematurity and convergence out of optimized point for genetic algorithm, a fitness normalization formula is introduced. In simulated annealing algorithm, the metropolis rule is used to sample product design schemes. Experimental results show that our proposed algorithm is more time-efficient than exhaustion and standard genetic algorithm, and it can guarantee the diversity of the design schemes.