CASPIAN JOURNAL

MANAGEMENT AND HIGH TECHNOLOGIES

Definition of the name N of linguistic variable of output value of criterion component of physical principle of sensing element action

Read

Khomenko Tatyana V. Definition of the name N of linguistic variable of output value of criterion component of physical principle of sensing element action // Caspian journal : management and high technologies. — 2012. — №3. — pp. 83-89.

Khomenko Tatyana V. - Ph.D. (Engineering), Astrakhan State Technical University, 16 Tatishchev st., Astrakhan, 414025, Russia,, t_v_khomenko@ mail.ru.

At present the automated systems of exploratory design allow to calculate the output values of criterion component of synthesized physical principles of action (PPA) of sensing elements (SE) in the automatic mode. Traditionally, all the values of criterion components of physicоtechnical effects (PTE) included in structural diagram of PPA are presented by the experts in the form of accurate quantitative assessment. For expansion of possibilities of the automated systems of exploratory design it is proposed to use the values of criterion component, presented by experts in the form of qualitative assessment: a fuzzy/linguistic variable, which is considered to be a given one, if you have defined all the elements of tuple, including the name N of fuzzy/linguistic variable. In the result of calculations of the output values of criterion component there emerges a problem of definition of the name N of fuzzy/linguistic variable. This task is complicated by the fact that at the various types of the input values (qualitative – quantitative) as a result of operations, the membership functions of the resulting fuzzy sets, which are difficult to identify, can be obtained . For solution of the set task the feature of neural networks with adaptive resonance is used: to preserve the plasticity when you store new images and, at the same time, to prevent the modification of the old memory. It is shown that the parallel calculations of the neural network is faster and does not increase with the rise in the number of components of PPA. The development of the system, covering the tasks of calculation of criterion component of estimation of PPA of SE with clear and fuzzy/linguistic values and its implementation make it possible, first, to combine the various ways of presenting the values of criterion component of PTE estimation, secondly, to apply the developed system at the early stages of design of technical facilities.

Key words: Sensing elements of control system,Calculation of values of criterion component,Qualitative assessments,Fuzzy/linguistic variable,Name identification,Neural networks,Adaptive resonance