Research on the most integrated assembly oriented

With the rapid development of science and technology, the global market competition is becoming increasingly fierce, the variety of products, the complexity of products is increasing, and the delivery time of products is becoming shorter and shorter. According to statistics, the investment in product design is only 10%, but it determines 70% - 80% of the product cost, of which the assembly cost accounts for more than 40% of the manufacturing cost. Therefore, product assembly plays an increasingly important role in the process of product development. Designforassembly (DFA) is a quantitative analysis and Research on the assembly ability and assembly process of products in the early stage of product design, which can reduce the number of parts, improve the assembly process of products, reduce the assembly cost and assembly time of products, and solve various problems in assembly in the design stage. Therefore, it is of great significance to carry out assembly oriented design research

DFA technology, as a subset of concurrent engineering, was also widely used in weapon production in order to design layered composite structures in the early stage of World War II, and produced good benefits. However, the concepts of concurrent engineering, DFM and DFA were only put forward in the late 1980s and have developed rapidly. The DFA method is summarized as follows: Yamagiwa divides the assemblability into many factors for evaluation and analysis, and the assembly operation with low evaluation score is easy to identify and modify in the early stage. Warnecke analyzes and evaluates the ratio of function and assembly difficulty, and takes the parts with low function value and difficult assembly as the object of guiding design modification. Hitachi method uses 20 symbols to represent various assembly operations, each symbol represents an operation or an assembly relationship, and the evaluation score is set according to the difficulty of the assembly operation (calculated according to the percentage system). Lucas divides the parts into class A (parts with general design requirements) and class B (parts with special design requirements), and the design efficiency is a/(a+b). Class B parts are reduced through redesign. Li carries out assembly evaluation based on assembly features, and puts forward the integration idea of CAD and DFA. Sturges divides the assembly into three levels of evaluation: part level, system level and process level. Fazio evaluates the assembly difficulty according to the assembly feature information, that is, feature position, assembly direction, clearance between parts and other information. Steven evaluates the assembly according to the efficiency of manual assembly. In recent 20 years, the efficiency of manual assembly: EA, m=ta, I × Nmin, i/ta, e, where: TA, I is the ideal assembly time, NMI supplements the surface moisture n, I is the ideal minimum number of parts, Ta, e is the estimated assembly time [9]. Hsu carries out assembly evaluation based on assembly constraint network. Constraint types: tool constraint, geometric constraint, physical constraint, functional constraint, clamping constraint. Boothroyd reduces assembly costs by adding other design constraints. First, reduce the number of parts assembled, and then ensure that other parts are easy to assemble. These methods lack the direct influence of factors such as the number of parts and the assembly relationship between parts on the assembly ability of products as a whole. Further research is needed to evaluate the assembly ability of products from the overall structure of products, so as to guide the assembly design of products

1Basic concept

definition 1: the generalized key LG is an information set that describes all assembly relationships between two assembly objects (parts or subassemblies). The two assembly objects of each generalized key may contain one or more assembly features, that is, multiple reassembly relationships. The generalized key can be expressed as: LG (I) = {r (1), R (2),..., R (n)}, which represents all assembly relationships between two assembly objects

definition 2: the strength of generalized bond fuzzy connection is a fuzzy measure of the assembly difficulty of generalized bond between two nodes (parts). It is a fuzzy synthesis of the connection strength of the assembly features in the generalized action, which is more compact and fast, and the system operation is more reliable and efficient. The strength of the assembly connection between two nodes is a measure of the difficulty of node assembly

definition 3: fuzzy graph is a triple

(1)

, where p={p1, P2,..., pn} is a set composed of N nodes (called universe); Is a fuzzy set on the universe P, and its membership function is μ I (PI), i=1, 2,..., N, which represents the fuzziness (or existence) of node PI. Here, the fuzziness of node is defined as the fuzzy assembly difficulty of factors related to the part itself; Is domain p × The fuzzy relation on P can be expressed as matrix

(2)

, where μ ij(i=1，2，…，n,j=1，2，…，n)； 0≤ μ ij≤1。 When μ If ij ≠ 0, there is an edge between nodes, μ Ij is called the connection strength between nodes. In this paper, the connection strength between nodes is defined as the fuzzy assembly difficulty of the edge (assembly feature). μ Ij=0 means there is no edge connection between them, that is, there is no assembly relationship: μ Ij=1 means that the edges are connected, that is, there is an assembly relationship, and the assembly difficulty is 1, 0 ≤ μ Ij ≤ 1 indicates that there is a fuzzy assembly difficulty between them μ Fuzzy edges of ij. μ ij= μ Ji is symmetric. If a part of a fuzzy graph still constitutes a fuzzy graph, the graph is a fuzzy subgraph of the original graph. Edges in a directed fuzzy graph have directions, so