Choosing optimal pattern matching algorithm for designing production system
Оксана Василівна Захаркевич, Анастасія Василівна Почупрін · Eastern-European Journal of Enterprise Technologies · 2014
The main purpose of this study is to formalize the principles of representing the features of the production system and antecedents of productions for implementing the optimal pattern matching algorithm. Optimality of this choice determines the production system efficiency since the problem of pattern matching is the most memory and time-consuming. When analyzing the experimental studies of the existing pattern matching algorithms, the main problem is that they involve different performance indicators, i.e. the characteristics for comparison may be selected so that they give preference to one of the algorithms in advance. Sample of universal characteristics of production system, taking into account the features of the solvable application task, including the properties of the antecedents of productions is proposed in the research. Based on the values of these characteristics, the estimate of each of the existing pattern matching algorithms can be calculated using the proposed schemes. The maximum value corresponds to the preferred algorithm. This choice, in turn, allows to determine or modify existing software tools for implementing computational formalism, based on the production model. An example of using the methodology for implementing a software module to define the technical state of NPP facilities is given. The research results can be used in designing the production system, aimed at solving the pre-formalized application task, mainly of technical and medical diagnostics