Expert system in technological preparation of parts assembling process on printed circuit board

Аuthors

Dembitsky N. L.^*, Lutsenko A. V.¹, Pham V. A.^2**

1. Lianozovskiy Electrical and Mechanical Plant, 110, Dmitrovskoe shosse, Moscow, 127411, Russia
2. Moscow Aviation Institute (National Research University), 4, Volokolamskoe shosse, Moscow, А-80, GSP-3, 125993, Russia

*e-mail: ndembitsky@gmail.com
**e-mail: phamvietanhqs@gmail.com

Abstract

The article considers the approach to solution of the problem of production method options selection applied in assembling electrical components on a printed circuit board (PCB) implementing a knowledge base and expert system. The study aims at creating the software providing process design allowing automatize the primary information processing on the parameters of production methods options realization.

Introduction of expert system to the software was made necessary due to the need to analyze interrelated factors, affecting the effectiveness of a technological process. All the factors were combined into eight groups with in accordance calculating assignments. These groups of factors generate libraries of basic program classes, which encapsulate inside the main features and methods of production methods parameters processing. The interaction of these classes is realized through the external memory (blackboard), where the values of all parameters of a production method are stored.

The study considers the example of implementation of the suggested model for estimation of selection of the equipment for radio components on PCBs according to the effective items peak output criterion. The parametric model of assembly errors contains four logically interconnected models: component model, PCB model, assembly robot model, and technological operation model. Each of the models determines analytical and logical dependencies of the factors affecting the effectiveness of a production method, and is represented in the program complex by the class.

The developed mathematical models are generic models for prediction of production method of PCB-based units’ efficiency. They allow perform calculations of quality evaluation and production method options selection according to reference books data (components structure, equipment parameters and materials)

The knowledge base in the form of library of classes provides opportunities to expand the functionality of the software for optimization of technological preparation for assembling PCB-based units. The output of expert system provides optimization problem model in canonical form, which is used in the optimization algorithm

The suggested approach is experimentally and small-scale production oriented, where the performance of highly costed analysis is limited by financial resources and the lack of qualified specialists.

Keywords:

the expert system, knowledge base, technological process, software complex, equipment, quality criterion, constraint, airborne radio equipment

References

1. Votintsev A., Borisenkov S. Elektronika: Nauka, Tekhnologiya, Biznes, 2005, no. 1, pp. 48-55.

2. Dembitskiy N.L, Lutsenko A.V., Pham Viet Anh Elektronnyi zhurnal «Trudy MAI», 2015, no. 81, available: http://www.mai.ru/science/trudy/eng/published.php?ID=57879 (accessed 13.05.2015).

3. Pham Viet Anh. Tezisy dokladov 13-i Mezhdunarodnoi konferentsii «Aviatsiya i Kosmonavtika — 2014», Moscow, 2014, pp. 468-469.

4. Dembitskiy N.L., Nazarov A.V. Primenenie metodov iskusstvennogo intellekta v proektirovanii i proizvodstve RTU (The use of artificial intelligence methods in the design and manufacturing of radio-equipment), Moscow, MAI, 2009, 228 p.

5. Dzharratano Dzh., Rayli G. Ekspertnye sistemy: printsipy i programmirovanie (Expert Systems: Principles and Programming), Moscow, ID Williams, 2007, 1152 p.

6. Frolov V.N., L’vovich YA.Ye., Metkin N.P. Avtomatizirovannoe proektirovanie tekhnologicheskikh protsessov i sistem proizvodstva RES (Computer-aided design of technological processes and production systems of REF), Moscow, Vysshaya shkola, 1991, 463 p.

Download