Search

Automated synthesis of computer-aided assembly planning (CAAP) processes is a crucial task for engineering practice and design theory. It is of especial relevance for modern robotic industries, which need in technological assembly instructions to be described in-depth and in-detail as much as possible. The assembly sequence is a key design decision on which many operational properties of the product and economic characteristics of production depend.Choosing a rational assembly sequence planning (ASP) is a challenge. It requires significant computing resources and taking into consideration a large number of technical parameters and economic characteristics that affect the quality of design alternatives. Insights into the quality of alternatives are given as the expert’s preferences rather than as the numerical criteria.The abovementioned features do not allow us to apply the classical optimization methods or mathematical programming for making ASP decision. For this, most modern publications offer various search engine optimization methods based on biological and behavioral analogies. In this paradigm, it is believed that a set of acceptable alternatives that form the original choice space is a priori known. In most design situations this presumption is unrealistic.In engineering practice, considerable technological knowledge about the assembly of products for different function purposes is gained. These are mostly ad hoc data that exist in the form of rules, recommendations, recipes, heuristics, expert preferences, descriptions of successful precedents, etc. The paper suggests a new method for a choice of the rational assembly sequence based on the use of the decision theory apparatus. The proposal contains formalization of important design and technological heuristics, namely consistency with the dimensional chain system, geometric “freedom” during assembly, monotony in size, weight, accuracy, etc.A set of choice functions is open and can be completed by additional choice functions that describe engineering heuristics and decision rules that are relevant in the given design situation. The proposed approach allows the assessment and choice of alternatives according to several aspects or criteria. To do this, it is possible to use various methods of generating a common choice function from the totality of particular functions.