Research On Characteristics Of Mechanical Self Organization Based On CA And Assembly Network

Mechanical design is the foundation of mechanical manufacturing. The cost of mechanical design is less than other stages, but the design process decides the cost and characteristics of mechanical product. The improvement of design quality can reduce production cost and improve production efficiency. Therefore studying on the mechanical design method can promote the development of machinery manufacturing industry. Computer aided design (CAD) largely simplifies the design work of designers, but its role is only to concrete and model mechanical products designed by designers. It is difficult to assist the design process which is more rely on the designer’s experience and knowledge. It is important to mechanical design to explore a suitable mechanical design automation method and to let computer aid or even replace designers in mechanical design. Mechanical self organization the design process of what the mechanical system evolves to mechanical products by using the system internal elements and self evolution rules without external intervention. It is a research direction of mechanical automation design.This paper mainly use cellular automata theory and assembly network theory to study mechanical products self organization process and characteristics. Starting with theory of complex network, the study optimized the assembly network theory according to the characteristics of small world networks, scale-free networks and social networks. And it gave more accurate description of initialization conditions of assembly network model. And then this paper studied on assembly network growth mechanism. The cellular and cellular space of assembly network was set according to cellular automata theory. The cellular state was set based on classification method of mechanical parts. The cellular automata local evolution mechanism was studied according to the characteristics of real network parts. Through the research of two aspects above, assembly network cellular automata model is established. The experiment simulated the growth process of assembly network and verified the effectiveness of the model. At last self-organization characteristic of assembly network was summarized by the comparison and analysis of simulation results.