Modularity Analysis Of Complex Products For Industry 4.0 And Its Intelligent Optimization Algorithm Design

To meet massive and customized demands in industrial environments proposed in the Industry 4.0 strategy,it is essential for enterprises reduce production costs and shrink design time in a highly competitive market.The traditional way of designing products cannot longer satisfy the fluctuate demands.However,through combining products' structure with design process,modular design is capable of massive production with various demands.Specifically,modular design is a way of dividing parts of a product into groups according to their different characteristics,like function.Therefore,it is meaningful for future changes in products' structure or new products to come up with a standardized modular design.As a new evolutionary algorithms,swarm intelligence algorithms can find near-optimal solutions by cooperation between individuals.Without centralized control and full knowledge of all possible solutions,swarm intelligence algorithms find solutions for complex distributed problems in a reasonable time.Owing to conciseness in concept simplicity in realization,they become focuses in most of recent studies.By researching into recent studies,an algorithm is proposed in this paper to solve a modular design problem of complex products in manufacturing industry.Firstly,modular design of complex products in the industrial environments is introduced.A mathematic model is built and it is proved to be NP-Complete on the basis of graph theory and computational complexity theory.Secondly,a hybrid of genetic algorithm and ant colony algorithm(HGACO)is proposed in this paper.Genetic algorithms(GA)is employed to obtain visibility distribution and ant colony algorithms(ACO)is then used to generate solutions.Also,operators like crossover and mutation are adopted to construct a new individual.HGACO has an advantage over the traditional GA for there are no longer large number of redundant solutions in the later phase of evolutionary process.Besides,computational experiments show that HGACO has superiority over others in both optimization performance and computational efficiency.The main contribution of this paper lies in: 1)considering massive and customized demands in the Industry 4.0 strategy,a model is built for a modular design problem of complex products;2)modular design of complex products is proved to be NP-Complete;3)HGACO is proposed to obtain solutions with good quality and low computational costs.